This document discusses economic and financial mechanisms for reducing disaster risk, including insurance and microfinance. It notes that poverty can increase vulnerability to hazards, so economic development and poverty reduction are important for risk reduction strategies. Livelihood diversification is key for poor people to reduce vulnerability by having multiple sources of income. However, development programs need to consider hazard risks so assets are protected. Insurance is an effective risk-sharing method, but is market-driven so mainly benefits those in developed countries where risks can be accurately calculated. Microfinance, like loans, can help build assets but borrowers risk losing assets if a disaster strikes before returns are generated.

1. Good Practice 4th
8/3/04
2:44 pm
Page 213
Chapter 13
Economic and financial mechanisms
for risk reduction
13.1 Economic mechanisms
Economics is a central issue in disaster reduction because poverty can make
people vulnerable to many different hazards, and certain patterns of socioeconomic development can expose people to greater risks. It follows that
appropriate economic development and the reduction of poverty are essential elements of any risk reduction strategy.
13.1.1 Livelihood diversification
Economic diversification is central to poor people’s strategies for reducing
their vulnerability to external shocks. They seek to do this primarily by
increasing their sources of income, building up a strong and diversified asset
base, managing their money well and maintaining access to multiple sources
of credit. This increases income overall and reduces dependency on individual sources: such dependency is a major contributor to vulnerability.1
213
13
Possession of wealth and assets gives households a wider range of options in
times of crisis, and speeds their recovery from disasters (see Case Study
13.1). Many disaster mitigation programmes pay no attention to this, considering it a task for development agencies. The mitigation programmes implemented by members of the Citizens’ Disaster Response Network in the
Philippines are an exception. They address a wide range of livelihoods issues,
focusing on food security and nutrition and including: crop and livelihood
diversification, propagation of disaster-resistant crops, supporting seed
banks and plant nurseries, improving post-harvest facilities, encouraging
better land-use management and sustainable agricultural practices, training
community health workers, establishing village pharmacies and medicinal
herb gardens, holding literacy classes and improving paths and footbridges.2
economic
and financial
mechanisms
Many coping strategies are based on this principle. Multi-cropping improves
farmers’ chances of being able to produce some food for consumption and
sale should the rains fail. Male migration from the countryside to towns to
seek labouring work during droughts provides an alternative source of
income to compensate for falling crop production.

2. Good Practice 4th
8/3/04
2:44 pm
Page 214
disaster risk reduction: mitigation and preparedness
Case Study 13.1
Poverty and disaster: a cyclone in India
A wealthy and a poor family live 100
metres apart near the coast of
Andhra Pradesh in south-east India.
The wealthy family has six members,
a brick house, six cattle and three
acres of land. The head of the
household owns a small grain
business and a truck. The poor
family – husband, wife and two
children – has a thatch-and-pole
house, an ox and calf, half an acre of
poor land, and sharecropping rights
to another quarter of an acre.
When the cyclone strikes, the
wealthy farmer has received a
warning on his radio and leaves the
area with his family and valuables in
the truck. The storm surge (flood)
brought by the cyclone partly
destroys his house and the roof is
taken off by the wind. Three of his
cattle are drowned and his fields are
flooded, destroying the crops. The
youngest child of the poor family is
drowned. The family’s house is
destroyed. Both animals are
drowned. The fields are flooded and
the crops ruined.
The wealthy family uses its savings to
rebuild the house within a week. The
lost cattle are replaced, and the fields
ploughed and replanted. The poor
family does not have savings and has
to borrow money for essential shelter
from a local money-lender, at
exorbitant rates of interest. It manages to buy a calf, but has to hire
bullocks to plough the field – which is
done late, since many others are in
the same position and draught animals are in short supply. As a result, the
poor family goes through a hungry
period eight months after the cyclone.
P. Blaikie, T. Cannon, I. Davis and B.
Wisner, At Risk: Natural Hazards, People’s
Vulnerability and Disasters (London:
Routledge, 1995), p. 47.
13.1.2 Protecting assets
‘Pro-poor’ development is not necessarily synonymous with vulnerability
reduction, and greater wealth may not by itself reduce risk.
The Kobe earthquake in 1995 is a good illustration of this point. Although
Japan is a prosperous, technologically advanced country, which has invested
heavily in mitigation and preparedness, there was a degree of complacency in
government and society about the level of preparedness and safety, and
disaster planning and management systems were inadequate. Many of the
214

3. Good Practice 4th
8/3/04
2:44 pm
Page 215
chapter 13 economic and financial mechanisms for risk reduction
5,500 deaths and much of the more than $100bn in economic losses could
have been avoided.3
Assets do not protect themselves against hazards: they have to be protected.
Any project seeking to build up vulnerable people’s livelihoods must take the
hazard context into account.
When poor people borrow money to buy livestock, tools or raw materials that
can be used for income-generating activities, they are increasing their livelihood assets, which, in the long run, will help them to become more resilient to
many external shocks. But drought, floods or other hazards can wipe out these
assets before they have been able to generate much return on the initial investment. In such cases, those concerned actually become worse off: not only are
they without assets, as before, but they also have a loan to pay back.
Similar scenarios can be drawn for other development interventions – investment in a new house or workshop where there is a risk from severe flooding,
for example, or construction of concrete irrigation channels across an
unstable hillside. In some cases, poor and vulnerable people may make a
conscious decision not to invest too heavily in particular items, choosing
those that can be replaced cheaply and easily. This is an important factor in
technology choice (see Chapter 10).
Economic development and poverty alleviation programmes need to take the
hazard context into account if they are to have a meaningful impact on
vulnerability reduction in the long run.
13.2.1 Insurance
Insurance is a standard and effective method of sharing risk, especially in
developed countries, where it has been in use for over 300 years.
Nature and benefits of insurance
In theory, everyone benefits from insurance. Individuals and organisations
buy it so that they can be compensated when hazards lead to death, injury or
ill-health, and loss of property or income. Claims are paid quickly and without
215
13
Two kinds of financial instrument will be discussed in this section: insurance
and other forms of micro-finance (especially micro-credit).
economic
and financial
mechanisms
13.2 Financial mechanisms

4. Good Practice 4th
8/3/04
2:44 pm
Page 216
disaster risk reduction: mitigation and preparedness
conditions attached. This gives policy-holders the promise of some financial
stability, and hence the confidence to invest (e.g. in home improvement) or
expand (e.g. a business enterprise). Insurance companies protect themselves
against major catastrophes by basing insurance premiums on sophisticated
calculations, spreading their own exposure across many different areas and
types of risk, and buying their own insurance cover (reinsurance).
Although the industry operates in a volatile environment, has suffered severe
blows in developed countries – most famously from Hurricane Andrew in
Florida in 1992, which forced several US insurers out of business – and is
becoming very concerned about the likely impact of climate change, it has
until now proved fairly robust, even when faced with major crises. Insurance
funds transferred rapidly to the US following the terrorist attacks of 11
September 2001 helped to reduce the country’s balance of payments deficit
and gave the economy some stability when it was urgently needed.4
Insurance is predominantly commercial. Decisions about whether to buy and
sell insurance, what kinds of cover to provide and what premiums to set, are
determined by market forces. Governments sometimes intervene, either
through state schemes (to protect farmers against crop losses, for example)
or by making some kinds of insurance cover compulsory (such as employers’
liability or motor insurance). Other institutions can also force people to take
out insurance: companies making loans to people to buy houses usually
insist that they are insured.
Insurance can stimulate other risk reduction measures. Insurers may only
provide cover in high-risk areas if governments ensure adequate mitigation
measures and emergency management systems. State crop insurance
schemes may allow farmers to take the risk of planting different crops,
leading to greater diversification and security against individual hazards. In
the US, the National Flood Insurance Program is a partnership between
public and private sectors that links the premiums paid to the level of protection: once government emergency management specialists have certified
that communities and households have put particular mitigation measures in
place, they are eligible for lower rates from the commercial insurers involved
in the scheme.5 In Fiji, houses must be certified by a structural engineer that
they have features making them cyclone-resistant before they can secure
cyclone insurance cover and mortgages.6
Because of the size of the global insurance industry and its obvious value in
reducing risk through risk-sharing and stimulating mitigation, disaster
agencies have become interested in its potential for protecting the most
216

5. Good Practice 4th
8/3/04
2:44 pm
Page 217
chapter 13 economic and financial mechanisms for risk reduction
vulnerable. However, because the industry is market-driven, its success is
based on confining its coverage to places where the risks can be calculated
with some accuracy and certain minimum standards such as building codes
can be enforced, and to people who can afford to pay. In effect, this means
that coverage is limited to developed countries and to wealthy people and
larger businesses in developing countries. Commercial insurance is unavailable to the people who arguably need it most: the poorest and most vulnerable in developing countries. Insurance companies have shown hardly any
interest in extending their coverage to such groups.
The market impact of major disasters has driven companies to raise
premiums substantially, or even withdraw cover in high-risk areas such as the
hurricane-prone Caribbean and south-east US. State insurance is available in
many developing countries, but premiums tend to be high and policies are
aimed at the professional classes in urban areas.
Insuring the poor against disasters
There has been little attempt to develop wholly commercial insurance
programmes targeted at poor and vulnerable people. Experience so far is not
encouraging (e.g. Case Study 13.2). Some experts question if it is even
possible to provide insurance cover to the poor on a commercial basis.
Where insurance schemes for poor groups and individuals have been
successful, they have generally originated in development programmes that
217
13
Initiatives to develop more flexible insurance tools to support poor and
middle-income countries are at an early stage; many work at national rather
than local levels, and they do not necessarily encourage greater emphasis on
measures to reduce the impact of future disasters.
economic
and financial
mechanisms
Even in wealthy societies, many people may not take out insurance. Only 3%
of homeowners in the areas hit by the 1995 Kobe earthquake in Japan had
insurance cover: 147,000 affected homes were uninsured.7 In such circumstances, the burden of financing recovery is passed on to governments. Many
people eligible for reduced premiums through the US National Flood
Insurance Program do not take out insurance cover, believing that in the
event of a disaster the national government will be obliged to compensate
them for their losses anyway.8 All insurance schemes have to face the
problem of ‘moral hazard’, where the sense of security and confidence that
insurance cover provides leads to people failing to take steps to reduce risk,
or even to take greater risks.

6. Good Practice 4th
8/3/04
2:44 pm
Page 218
disaster risk reduction: mitigation and preparedness
Case Study 13.2
Developing disaster insurance in Cambodia
In 2000, Cambodia experienced its
worst floods for 70 years. More than
three million people were affected,
mostly poor farmers; half a million
were displaced and nearly 400 killed.
Direct costs were estimated at
$160m – equivalent to half the
government’s recurrent budget for
that year.
In consultation with the government,
international agencies, the private
sector and communities, a World
Bank team drew up a proposal for a
project to investigate how private
insurance might be used to provide
relief to poor farmers. The project
documented current and historic
flood data, climatic conditions and
rice yields, and looked at ways of
collecting and managing premiums,
assessing claims and giving out
payments. Surveys were also carried
out to determine the impact of past
floods on the poor, how farmers had
coped, their capacity to save and
how they used savings. Farmers
were prepared to save, albeit only
very small amounts, and
communities welcomed the idea of
insurance.
However, commercial reinsurers
were unwilling to take on the risk of
insuring poor communities likely to
be affected by floods every two or
218
three years. It was found that
damage assessments would be
costly and time-consuming, which
would delay compensation
payments. There was a risk of
farmers having no incentive to take
measures to protect themselves
because they would expect the
government or insurers to cover
their losses anyway. To make
premiums affordable and spread
the risk would have required
widespread participation by
farmers, and it was felt that this
might not be possible without some
compulsion.
The project therefore looked at
alternative ways of spreading the
risk. These included governmentsupported insurance pools, local
catastrophe funds, and building on
local savings and credit schemes or
informal mechanisms for selfinsurance.
World Bank, Cambodia Disaster Insurance
(Project 1071): First Interim Progress
Report, 21 September, 2000; Cambodia
Disaster Insurance (Project 1071): Second
Interim Progress Report, 16 February,
2001; Working to Develop Disaster
Insurance in Cambodia, 2001–2002,
www.developmentmarketplace.org/report
1071.html, www.worldbank.org/
developmentnews/stories/html/
010802a.htm.

7. Good Practice 4th
8/3/04
2:44 pm
Page 219
chapter 13 economic and financial mechanisms for risk reduction
have aimed at financial sustainability rather than profit. Although there is still
a lot to learn about how to make such schemes work effectively, and many of
them are fairly new, recent research on the subject has documented some of
these experiences in detail and provides valuable lessons.9
Such schemes are run mainly by micro-finance institutions but also by NGOs,
cooperatives, governments and even companies. Business involvement is
often in partnership with non-profit organisations, where the business –
usually an insurance company – typically provides technical expertise (e.g.
actuarial calculations regarding risks and their likely costs), assistance with
marketing, or underwriting. Some schemes have an outreach of a few
hundred families, but they can reach large numbers of people – millions, in a
few cases. Terms and conditions vary widely, as do the administrative and
financial structures used, but life insurance, which is the main form of insurance on offer, tends to be compulsory: people who wish to borrow money or
open a savings account with the organisation concerned have to buy an
insurance policy. The other main kinds of coverage are health and, to a lesser
extent, property insurance.
As this is still a new field, lessons are being learnt all the time. Schemes may
have to be amended repeatedly to achieve the right balance between effec-
219
13
The insurers have to put considerable effort into marketing their schemes to
people who are poor and unfamiliar with the concept and workings of insurance. One health insurance programme in Colombia claimed to have enrolled
16,000 members in seven months, but others have encountered considerable
resistance.10 Local-level approaches involving community meetings and
regular discussions with field workers appear to be most successful. Once
households understand insurance, interest seems to be strong. Where there
is already a relationship of trust between the insurer and the community –
notably where the insurer is an established micro-finance institution or NGO –
a base of policy-holders can be built up quickly.
economic
and financial
mechanisms
The evidence available indicates that insurance programmes for poor people,
especially life insurance, can be financially viable, though the impact of major
disasters on programmes has yet to be studied. In any case, insurance is a
risky business. To maintain financial stability, life policies generally exclude
high-risk groups such as the elderly, and certain causes of death such as
epidemics. Health insurance may exclude health care costs for AIDS-related
treatment or injuries arising from involvement in riots or other civil unrest.
All-risk coverage for property has largely been abandoned after unsuccessful
attempts in the 1970s and 1980s to provide all-risk crop insurance to farmers.

8. Good Practice 4th
8/3/04
2:44 pm
Page 220
disaster risk reduction: mitigation and preparedness
tive coverage of poor and vulnerable people on the one hand, and financial
sustainability on the other. This is particularly true of health insurance
services, whose financial performance is much lower than that of life insurance. It may be advisable to take a long-term approach to sustainability.
A major challenge is to find commercial partners who are willing to work with
low-income people. This is a particular problem in reinsurance (as in Case
Study 13.2 above), although a few non-profit agencies have been able to
obtain it. The availability of reinsurance is the main factor affecting market
growth throughout the insurance industry.
Another problem is delays in settling claims. This is partly the result of
bureaucratic slowness, but partly inevitable where claimants live in remote
villages or communications break down because of technical failures, environmental hazards or civil unrest.
There are examples of successful mutual benefit societies, where insurance
funds are set up by groups to provide cover for their members. For example,
in Mexico there are 200 mutual insurance funds (fondos) for farmers, with
nearly 70,000 members. Each member pays an annual premium, which
provides cover against crop failure due to drought, flooding, pests, disease or
other emergencies. Surplus funds at the end of the year are used to provide
services to members. The government provides technical support and
subsidises the premiums, but the fondos have performed well financially and
the intention is to raise private finance in due course. However, admission to
membership of a fondo depends on its perception of a farmer’s capacity, and
the scheme has been criticised for excluding poorer, more vulnerable
farmers.11
Micro-insurance can be used to stimulate mitigation activities. Health insurance is often linked to preventive and primary health care programmes run by
the insurer concerned (if an NGO) or a partner organisation, and policyholders may be expected to use such services. A scheme in Colombia offered
regular group meetings for policy-holders to teach them how to protect their
property against common hazards.12 SEWA (see Case Study 13.3) provides
fruit and vegetable vendors with umbrellas, having concluded from its life
insurance data that spending long hours sitting in the sun might be
increasing the risk of illness.13
As well as running insurance schemes, NGOs provide technical support to
schemes run by other agencies. Médecins sans Frontières, for example, has
provided a doctor as part of a rural health insurance scheme in Cambodia.14
220

9. Good Practice 4th
8/3/04
2:44 pm
Page 221
chapter 13 economic and financial mechanisms for risk reduction
Case Study 13.3
Insuring the vulnerable
An annual premium of Rs75–360
provides different levels of life
insurance and coverage for the costs
of health care and property losses
The two insurance company partners
provide the coverage for life, health
and asset insurance (life insurance is
partly subsidised by the
government). Maternity benefits and
administrative costs are paid out of a
fund established by GTZ. Another
GTZ grant supports the programme’s
training and educational work.
By 2002, the insurance scheme had
paid out claims totalling over Rs14m
to more than 10,000 women.
S. Sinha, Strength in Solidarity: Insurance
for Women Workers in the Informal
Economy (Ahmedabad: Self Employed
Women’s Association, 2002).
Non-profit and community organisations could also play a part in lobbying
governments and others to establish non-profit insurance schemes or to
create the conditions that will encourage the spread of commercial insurance
cover – if not for the poor, then at least for lifeline facilities such as hospitals,
schools, power plants and bridges.
Informal insurance and social security systems are widespread among the
poor in disaster-prone societies, though mostly in rural areas. Borrowing and
221
economic
and financial
mechanisms
SEWA’s integrated security insurance
scheme, which began in 1992, covers
70,000 members. It is run in
collaboration with two national
insurance companies and the
German government’s aid agency
GTZ, and is administered by SEWA
Bank, which manages the
Association’s extensive savings and
credit programme.
arising from floods, fires and
communal riots. A small maternity
benefit is also available. An alternative
payment is by fixed deposit, the
annual interest from the deposit being
credited as the annual premium.
13
In western India, the Self Employed
Women’s Association (SEWA), with a
membership of over 220,000 poor
women, provides a range of services
including organising members into
trade organisations and
cooperatives, providing loans,
training and technical assistance,
legal aid, child care and education.

10. Good Practice 4th
8/3/04
2:44 pm
Page 222
disaster risk reduction: mitigation and preparedness
sharing are part of the social fabric in good times and bad. They comprise an
important coping strategy during crises, when those who are suffering can
call on neighbours or kinsfolk for food, materials or other support. Exchanges
of food form an important part of famine mitigation strategies in SubSaharan Africa. Funeral or burial societies are found throughout the world,
their members pooling funds to cover expenses related to the death of
another member.
Custom on some of the chars (silt islands) in the Jamuna River in Bangladesh
allows people who have lost their homes to erosion by the river to rebuild on
anyone’s land. It is understood that the favour will one day be returned: at
some stage during their lifetime, most if not all of the islands’ inhabitants will
be forced to move as the river changes course.15 In Uganda, self-help groups
known as Munno Mukabi (Friend in Need Associations) save money that can
be lent to members whenever a crisis strikes. Members also promise to make
their labour available to others during a crisis or for a celebration.16
13.2.2 Micro-credit and other forms of micro-finance
Micro-finance programmes are numerous and widespread in many developing countries, with a massive outreach numbering millions of people. Many
NGOs run savings and credit schemes, which often form an important
element in their development programmes. Such organisations are also
developing a wider range of financial services for poor people. Although
there are debates on the extent to which micro-finance contributes to poverty
reduction and its influence compared to other factors, it is generally acknowledged to play a significant role.
Micro-finance is also important in reducing vulnerability before disasters and
supporting post-disaster recovery. Its considerable strategic potential in
these areas is only now becoming understood and utilised. Organisations
that manage savings and credit programmes for the poor – usually referred to
as micro-finance institutions (MFIs) – should be more fully integrated in risk
reduction initiatives.
Research shows that loans, which are primarily invested in productive enterprises that generate income, are also often used to cope with present or
potential crises that threaten livelihoods – by laying in stocks of food, making
improvements to farmland, repairing houses, buying tools or other productive equipment, digging wells and irrigation systems, acquiring new skills, or
making gifts to family and friends so that reciprocal favours can be asked
later. After a disaster, credit is used by victims to speed recovery by replacing
222

11. Good Practice 4th
8/3/04
2:44 pm
Page 223
chapter 13 economic and financial mechanisms for risk reduction
lost assets and helping them get back to work. Loans are often taken out to
deal with household crises – especially those caused by sickness or death in
the family (which has both emotional and economic consequences), but also
by such shocks as food shortages, sudden price increases, loss of employment or theft.
Until recently, MFIs did not pay much attention to hazards and disasters. This
position altered rapidly in 1998, when flooding in Bangladesh and Hurricane
Mitch in Central America caused widespread death, injury and loss among
members of savings and credit programmes – and, as a result, damaged the
programmes themselves. In Bangladesh, more than 30% of MFI clients lost
their houses or moved to safe places; 65% suffered losses or damage to
business assets; and over 90% had to suspend income-generating activities
for more than three weeks. Loan recovery rates fell from 92% to 43%; MFI
staff could not locate borrowers or mobilise them for group meetings.17
Some post-disaster loans are made to replace or repair physical assets:
equipment for income-generating activities such as cooking utensils or
sewing machines; and rebuilding business premises. The objective of such
223
13
Some MFIs provide emergency loan facilities to their clients to meet immediate
needs for food, clean water or medicine. These too are made at lower interest
rates or even without interest, although it seems that many clients prefer to
borrow informally in such circumstances – from friends or relatives if they can,
and from money-lenders if they cannot. Even where savings and credit
programmes are available, informal borrowing remains important in poor
communities, especially if it is to be spent on consumption, rather than invested
in productive activity (at times of crisis, families go to great lengths not to use up
their savings or sell off their other livelihood assets). They are more likely to take
up emergency loans from MFIs if these can be made rapidly and come with few
or no restrictions on the purpose for which they can be used.
economic
and financial
mechanisms
Since then, MFIs have begun looking at how to protect themselves and their
clients against risk. A number of studies and good practice guidelines are
now available.18 Most interest to date has been in dealing with the consequences of disasters. MFIs use a variety of methods to help those who are
affected. Rescheduling loans has become a common practice. Writing off
loans is undesirable, because it undermines long-term commitment by clients
to repay as well as being a loss to the microfinance scheme itself. The terms
of rescheduling have to be varied according to a number of factors including
the nature and timing of the disaster, the community’s cash flow patterns and
the MFI’s own financial situation.

12. Good Practice 4th
8/3/04
2:44 pm
Page 224
disaster risk reduction: mitigation and preparedness
loans is to help households reconstruct their businesses and livelihoods. They
tend to be relatively large amounts and are usually made once the relief period
is over, at normal interest rates and with a longer repayment period. Only large
MFIs can afford to make a large number of asset replacement and housing
loans, and there is some evidence of higher than usual failures to repay. For
this reason, few MFIs are prepared to make such loans. Some feel that the
standard short-term working capital loan is best suited for disaster recovery.
MFIs need to react quickly in a disaster, in assessing the situation and
planning their response. This is not always easy, for communications may
break down. During the 1998 floods in Bangladesh, many local branches of
MFIs were inaccessible and it was not until the water receded that disaster
victims could withdraw savings, take out new loans or renegotiate old ones.
Branch staff need training to manage in such circumstances, crisis management plans should be drawn up, and clear guidance is needed on lending
policy and practice. MFIs should be linked to early-warning systems, and
ensure that their clients are informed about potential disasters.
In the absence of other relief agencies, MFIs often undertake short-term relief
work, especially if they are part of NGOs normally involved in other development and humanitarian activities. This can cause problems for staff who are
not trained as aid workers. MFIs engaged in relief also need to communicate
clearly to their clients that their efforts are temporary and one-off, and do not
influence their primary role as providers of finance.
MFIs can prepare for disasters in many ways. In Bangladesh, it is common
practice for MFIs to put a percentage of clients’ ‘compulsory savings’ into an
emergencies fund, which can be made available quickly to disaster-affected
borrowers in the form of emergency loans (compulsory savings are regular
deposits made by borrowers to build up collateral against their loans: normally
they cannot be withdrawn while loan repayments are still outstanding). The
Grameen Bank reported that 95% of its compulsory savings were withdrawn
during the 1998 flood, while two-thirds of the clients of another large lending
NGO withdrew more than half their compulsory savings. Such rapid, large-scale
withdrawals caused problems for smaller Bangladeshi MFIs which had reinvested compulsory savings in their standard loan programmes: some could
only release 25–50% of clients’ savings.19 The stricter the conditions attached
to the use of compulsory savings, the more likely it is that poor people will turn
to other sources of loans, including money-lenders.
MFIs may introduce preparedness and mitigation initiatives for their clients.
This is more likely where the micro-finance programme or institution is part of
224

13. Good Practice 4th
8/3/04
2:44 pm
Page 225
chapter 13 economic and financial mechanisms for risk reduction
a larger NGO’s portfolio. Some MFIs in Bangladesh have made subsidised
loans for emergency preparedness purchases such as food, fuel, water purification tablets and rehydration tablets. Housing loans may be provided in
normal times to help clients build in safer locations.
MFIs can encourage their clients to form insurance funds. In India, a federation of self-help groups called Vaigai Vattara Kalangiyam, encouraged by an
MFI called PRADHAN, operates its own welfare/disaster insurance fund
linked to the insurance scheme operated by the state insurance company and
similar to the SEWA scheme described in Case Study 13.3. In Mali and
Burkina Faso, MFIs have arranged for their clients to rent space in seed and
grain banks to store cereals as a reserve during droughts and famine. Some
MFIs insist that their members develop a contingency plan to deal with disasters. In Burkina Faso, this is reported to have been effective in reducing
arrears in loan repayment during drought in 1995.20
Like insurers, MFIs can spread their risk, by making sure that they serve poor
clients in areas less likely to be affected by hazards, or by lending to people
involved in more than one sector of the economy. For this, they need to
undertake their own risk assessments. But even well-prepared MFIs are
unlikely to be able to cope with disasters on the scale of those that
hit Bangladesh and Central America in 1998. For such events, additional
support is needed, which is why MFIs and international donors have begun to
experiment with donor-backed reserves that can be released quickly in an
emergency.
225
13
One important lesson from the 1998 Bangladesh floods was the need to
protect the MFIs themselves, which soon found themselves critically short of
funds as their disaster-affected borrowers failed to make repayments. The
practice of putting money aside into emergency reserves is becoming more
widespread in areas with regular disasters such as monsoon floods, but this
means that less money is available to lend to the poor on a regular basis, and
it is harder to justify in places where disasters are less frequent. An alternative is to make savings and credit groups pay a percentage of their own loans
into an emergency fund. In both cases, questions arise over ownership, rights
of access, decision-making, and terms and conditions.
economic
and financial
mechanisms
Evidence of MFI-stimulated disaster mitigation activity is patchy. It seems
that such initiatives have not been widely replicated. The potential appears to
be considerable, however, which highlights the need for more dialogue
between MFIs and other development and disaster management agencies
about collaborative efforts to reduce risk.

14. Good Practice 4th
8/3/04
2:44 pm
Page 226
disaster risk reduction: mitigation and preparedness
Case Study 13.4
Adapting micro-finance to disasters
Ashrai (the name means ‘shelter’ in
English) was set up in 1991 to
improve the social and economic
conditions of tribal people in the
Barind Tract in north-west
Bangladesh. It works with some
1,200 village societies representing
25,000 families. The main activity is
establishing and supporting savings
and credit groups, mostly women’s
groups. It also has its own revolving
loan fund, from which established
societies can borrow in equal
proportion to their savings.
From late 1997, a series of
emergencies led Ashrai to establish
and build up emergency loan funds.
The first was set up in January 1998
to help 5,773 families affected by a
severe cold spell to buy clothes and
blankets. The Swiss Agency for
Development and Cooperation (SDC)
made a grant to establish the fund.
Loans were to be repaid within 18
months, at an interest rate of 6% per
annum.
However, many of the families were
subsequently hit by the exceptional
floods later that year. The previous
emergency loans had to be
rescheduled, and 2,794 new
emergency loans were made to
repair houses and buy seeds, tools
and fertilisers. To do this, a second
emergency loan fund was
established with a second grant from
SDC, again at 6% interest per annum
but this time with a two-year
repayment period that reflected the
severity of the disaster. As in the
preceding winter, the local Ashrai
savings and credit societies and their
umbrella organisations helped to
identify the most needy.
In 1999, Ashrai and SDC decided to
merge the two emergency loan funds
into a single permanent fund for
future major events.
Because of the impact of the
disasters, people’s capacity to repay
the emergency loans was limited. By
December 2000, the emergency fund
had recovered 91% of the amount
disbursed because of extreme cold,
and 67% of its loans to flood victims.
Micro-finance in Disasters (London:
British Red Cross Society (London: NGO
Initiatives in Risk Reduction, Case Study
17, 2001), www.redcross.org.uk/riskreduction.
There is also potential for using non-financial credit – loans of seeds, tools or
materials – to help reduce risk. This does not appear to have been explored
226

15. Good Practice 4th
8/3/04
2:44 pm
Page 227
chapter 13 economic and financial mechanisms for risk reduction
widely, other than in the case of grain and seed banks (see Case Study 15.3,
page 264).
13.3 Chapter summary
• Economic development and poverty reduction are essential elements in
any risk reduction strategy.
• Livelihood diversification is central to poor people’s coping strategies, but
is generally overlooked in disaster mitigation programmes.
• Greater wealth reduces risk by giving households more options at times
of crisis and speeding their recovery from disasters. But livelihood assets
do not protect themselves: they must be protected against hazards.
• Insurance is a standard and effective method of spreading risk, but
commercial insurance is confined to richer countries and people, and
insurers are reluctant to cover poor, high-risk groups.
• Insurance schemes for the poor run by micro-finance institutions and
NGOs can be financially viable, but this is a new area of work and there is
still much to learn about how to make such schemes effective.
• Micro-finance has the potential to play a significant part in reducing
vulnerability before disaster, and supporting recovery.
• Most interest among micro-finance institutions has been in dealing with
the consequences of disasters (e.g. through rescheduling loans or making
emergency loans), although some work has been done to stimulate predisaster measures (e.g. creation of emergency funds, lending for specific
disaster preparedness activities).
2
3
4
5
6
7
J. Sebstad and M. Cohen, Microfinance, Risk Management, and Poverty (Washington DC:
Management Systems International (AIMS Project), 2000), pp. 48–75,
www.dec.org/pdf_docs/PNACJ418.pdf.
A. Heijmans and L. P. Victoria, Citizenry-Based & Development-Oriented Disaster Response:
Experiences and Practices in Disaster Management of the Citizens’ Disaster Response Network
in the Philippines (Quezon City: Center for Disaster Preparedness, 2001), pp. 57–63.
World Disasters Report 1996 (Oxford: Oxford University Press/IFRC, 1996), pp: 65–75.
E. Adamsdale, Transferring Risk: Potential Partnerships between the Insurance Industry and
the Humanitarian Sector (London: British Red Cross Society, 2002).
See the NFIP website: www.fema.gov/nfip.
C. Benson, The Economic Impact of Natural Disasters in Fiji (London: ODI (Working Paper 97),
1997), p. 70.
J. R. Heywood, ‘Natural Hazards as Problems for Insurers’, in Preventing ‘Natural’ Disasters –
The Role of Risk Control and Insurance (London: UK National Coordination Committee for the
International Decade for Natural Disaster Reduction, 1995), p. 16.
227
13
1
economic
and financial
mechanisms
Notes

16. Good Practice 4th
8/3/04
2:44 pm
Page 228
disaster risk reduction: mitigation and preparedness
8 D. R. Godschalk et al., Natural Hazard Mitigation: Recasting Disaster Policy and Planning
(Washington, DC: Island Press, 1999), pp. 68-9.
9 W. Brown and C. F. Churchill, Insurance Provision in Low-Income Communities, Part I: A Primer
on Insurance Principles and Products (Washington DC: Development Alternatives
(Microenterprise Best Practices Project), 1999), www.mip.org/pubs/mbp_def.htm; W. Brown
and C. F. Churchill, Insurance Provision in Low-Income Communities, Part II: Initial Lessons
from Micro-Insurance Experiments for the Poor (Washington DC: Development Alternatives
(Microenterprise Best Practices Project), 2000), www.mip.org/pubs/mbp_def.htm; Adamsdale,
Transferring Risk.
10 Brown and Churchill, Insurance Provision in Low-Income Communities, Part II, p. 49.
11 Adamsdale, Transferring Risk.
12 Brown and Churchill, Insurance Provision in Low-Income Communities, Part II, pp. 66, 70.
13 Ibid., p. 89.
14 Adamsdale, Transferring Risk.
15 H. Schmuck, Living with the Floods: Survival Strategies of Char-Dwellers in Bangladesh (Berlin:
ASA-Programm of the Carl-Duisberg-Gesellschaft, 1996), pp. 70–71.
16 Sebstad and Cohen, Microfinance, Risk Management, and Poverty, p. 58.
17 Rapid-Onset Natural Disaster Technical Briefs, Brief 1 (Washington DC: Development
Alternatives (Microenterprise Best Practices Project), undated),
www.mip.org/pubs/mbp_def.htm.
18 Ibid., Briefs 1–8; G. Nagarajan, Microfinance in the Wake of Natural Disasters: Challenges and
Opportunities (Washington DC: Development Alternatives (Microenterprise Best Practices
Project), 1998), www.mip.org/pubs/mbp_def.htm; J. Parker and G. Nagarajan, Can
Microfinance Meet the Poor’s Financial Needs in Times of Natural Disaster? (Washington DC:
Development Alternatives (Microenterprise Best Practices Project), 2000),
www.mip.org/pubs/mbp_def.htm; Sebstad and Cohen, Microfinance, Risk Management, and
Poverty.
19 Rapid-Onset Natural Disaster Technical Briefs, Brief 3.
20 Nagarajan, Microfinance in the Wake of Natural Disasters, pp. 22, 27.
228

17. Good Practice 4th
8/3/04
2:44 pm
Page 229
Chapter 14
Managing urban risk
14.1 Hazards, risk and vulnerability in urban areas
Half the world’s population lives in towns and cities. Within 30 years, this proportion may rise to two-thirds. The rate of urbanisation is greatest in the
South. About two-thirds of the world’s current urban population – 1.9bn people – live in low- and middle-income nations in Africa, Asia, Latin America and
the Caribbean. At least 600m of these are so poor that their income and other
assets are not enough to provide their essential needs.
The causes of urban growth are varied and complex, but among the main ones
are economic and environmental pressures driving poor people from the
countryside to seek a living in the towns. Poverty and poor urban management mean that many cities in the South are expanding rapidly and in an
uncontrolled way. This has led to severe social, economic and environmental
problems. It is also putting ever-greater numbers of people at risk from natural and technological hazards.
The themes discussed in this book apply just as much to urban as to other settings. At the same time, managing urban risks presents several distinct challenges. These arise from the specific features of urban hazards and vulnerability, and from the socio-economic and institutional environment in towns
and cities, which has a profound influence on the capacity of communities and
municipal authorities to address risk effectively.
The following discussion draws heavily on the work of the International
Institute for Environment and Development’s Human Settlements Programme
on urban environmental problems.1
• Biological hazards: diseases caused by unclean water and poor sanitation.
• Chemical hazards: polluted water, indoor and outdoor air pollution from
fires, chemicals, industrial processes and vehicles, dumped garbage and
hazardous wastes.
229
14
Environmental hazards are the main cause of ill-health, injury and premature
death in urban areas of the South. People living in towns and cities in developing countries face a large number of environmental hazards. They include:
managing
urban risk
14.1.1 Urban hazards

18. Good Practice 4th
8/3/04
2:44 pm
Page 230
disaster risk reduction: mitigation and preparedness
• Physical hazards: fire, floods, housing on slopes at risk of landslides,
earthquakes, cyclones, extreme temperatures, accidents due to inadequate infrastructure or facilities (e.g. open drains, missing manhole
covers, lack of pavements and places to cross roads).
Many of these hazards are also present in rural locations, but they become
particularly threatening in densely-populated urban areas. There is a strong
likelihood of major urban disasters in future, especially in mega-cities, many
of which are sited in earthquake zones or along coastlines struck by cyclones.
Earthquakes in major urban centres were responsible for some of the
greatest disasters in the 20th century – Tokyo in 1923, with 143,000 killed,
and Tangshan in China in 1976, where the death toll was probably 250,000.
The economic effects even of far less severe earthquakes can be colossal:
damages resulting from the 1995 Kobe earthquake amounted to over $100
billion.
Man-made and technological hazards such as fires, pollution, industrial and
road accidents assume much greater importance in towns and cities, where
there are also particularly acute public health problems arising from pollution
and inadequate sanitation in high-density settlements.
Air pollution presents particular problems for the elderly and children. This is
a largely urban hazard (apart from indoor cooking fires, which are a hazard
anywhere). Large numbers of people and vehicles and concentrations of
industrial activities produce considerable volumes of air pollutants. Where
these become trapped in the atmosphere, their concentration increases and
people are exposed to them for longer. This is a particular problem in highaltitude cities: Mexico City is notorious for its air pollution; in Kathmandu the
problem is growing rapidly. Air pollution is normally treated as a public health
issue rather than a problem for disaster managers, but its consequences can
be severe. An estimated 130,000 premature deaths and 50–70m incidents of
respiratory illness occur each year due to episodes of urban air pollution in
developing countries.2 In Santiago, Chile, a $911m, 15-year programme to
reduce severe air pollution from particulates, carbon monoxide and groundlevel ozone was launched in 1998. It aims to prevent 11,000 premature deaths
and 65m medical visits resulting from smogs in the city.3
Cities also tend to be much warmer than the surrounding countryside, and
night-time cooling is lower than in rural areas. This ‘heat island’ effect is caused
by the concentration of large heat-retaining structures, asphalt and concrete
landscapes, physical obstruction of cooling breezes by buildings, and heat
produced by industrial and domestic activities. In Mexico City, for instance,
230

19. Good Practice 4th
8/3/04
2:44 pm
Page 231
chapter 14 managing urban risk
temperatures can be 10°C higher than in surrounding areas.4 Many cities are
not well designed to deal with this problem. Ventilation and heat management
are rarely considered seriously in urban planning, or taken into account in
emergency planning, even in developed countries (see Case Study 14.1).
Case Study 14.1
Impact of an urban heat wave
231
S. A. Changnon et al., ‘Impacts and
Responses to the 1995 Heat Wave: A Call
to Action’, Bulletin of the American
Meteorological Society, vol. 77, no. 7,
1996, pp. 1497–1506.
managing
urban risk
Energy use for air conditioning and
other cooling increased rapidly,
leading to a power failure affecting
over 40,000 people. Highways and
railways were damaged by heaving
and buckling of joints and rails.
Many companies indicated that work
efficiency was greatly reduced.
The municipal authorities did not
respond well. Officials did not declare
a heat emergency – which would have
activated the city’s heat emergency
plan – until four days into the heat
wave. This was a major factor in the
high death count. Without advance
warning, few inhabitants used
Chicago’s five ‘cooling centers’: places
with air conditioning and other
facilities for people to relax and rest.
Thousands of heat-afflicted people
had to be taken to hospital during the
heat wave, but there were not enough
ambulances available and fire engines
had to be used. Hospitals could not
handle the high number of
admissions. In response to sharp
public criticism, the city’s mayor was
forced to appoint a commission to
discover what had gone wrong and
what should be done in future. This
led to a new, more comprehensive,
warning and emergency plan for
dealing with heat waves.
14
Heat waves claim an estimated 1,000
lives each year on average in the US.
A severe five-day heat wave in July
1995 caused 525 deaths in Chicago.
In this as in other heat waves, older
people were the prime victims: 73%
of those who died were people aged
65 or more. African-Americans were
nearly twice as likely to die from the
heat as whites. Most deaths were in
the inner city, where temperatures
did not fall off much at night, and in
homes without air conditioning or
where residents did not operate air
conditioning or fans because they
could not afford to. There were
suggestions that older people
without air conditioning were afraid
to leave doors and windows open at
night to cool their homes down
because of the perceived threat of
crime.

20. Good Practice 4th
8/3/04
2:44 pm
Page 232
disaster risk reduction: mitigation and preparedness
Urban food security presents a potentially serious risk, especially in megacities. The level of urban populations’ dependence on food brought in from
outside and purchased in markets is very high and will remain so, especially
in the largest cities. Serious problems can arise where food supplies are cut
or prices rise beyond the reach of the poor, and this may happen suddenly as
the result of political and socio-economic factors or natural hazards such as
droughts and floods.
Little attention has been paid to this risk in disaster management circles,
despite the obvious social and political consequences of food insecurity, the
difficulty of identifying vulnerable groups in this context and the considerable
operational challenge in obtaining large amounts of food and distributing it
to the most needy.5
14.1.2 Urban vulnerability
Poverty, gender, class, caste and ethnicity are powerful influences on urban
vulnerability, as they are in rural areas; but urban life adds new dimensions.
Poor and marginalised people are likely to live in poor-quality housing, in
neighbourhoods without clean water, drains and paved roads, where sanitation systems, garbage collection and public health services are inadequate.
This makes them vulnerable to a variety of environmental hazards.
Poverty forces people to live in the most polluted and dangerous areas: river
valleys, flood plains, hillsides, next to roads, waste dumps and hazardous
industries. Where they lack legal title to their property – as in many urban
settlements – and live in fear of eviction they have little incentive to invest in
private or communal mitigation measures, and in any case have little money
for doing so. Landlords are often unwilling to invest in their properties, and
may raise the rents of those improved by their tenants. Local governments
may refuse to provide services to informal settlements on the grounds that
this will imply recognition of the inhabitants’ right to the land where they
have settled.
High housing and population densities magnify the effects of pollution and
disease. Domestic fires, which are a significant risk in houses made of
materials that burn easily, such as wood, thatch and cardboard, can,
where homes are packed tightly together, easily get out of control.
Industrial accidents can be devastating. According to official figures, the
explosion at a chemical factory in the Indian city of Bhopal in 1984 resulted
in nearly 3,000 deaths and 100,000 injuries; more than 200,000 people
had to be evacuated.6
232

21. Good Practice 4th
8/3/04
2:44 pm
Page 233
chapter 14 managing urban risk
Case Study 14.2
Overcoming a food crisis in a mega-city
The Indonesian government’s
programme to provide food at
subsidised prices to those in need
The alternative supported by one
international donor was to use
existing commercial channels.
Indonesian firms ground imported
wheat into flour and used it to
produce noodles, which are an
important part of the urban diet.
These were sold through normal
outlets – retailers and pavement
stalls and restaurants frequented by
the poor – with prices fixed at levels
that were affordable but
nevertheless commercially viable.
Although it took time to get the
project going, it was able to reach
approximately a million people per
month over a period of eight months.
C. Kelly, Acute Food Insecurity in MegaCities: Issues and Assistance Options
(London: Benfield Hazard Research
Centre (Working Paper 7), 2003),
www.benfieldhrc.org.
Uncontrolled urban development damages the environment and increases risk.
Deforestation, land clearance and subsequent construction may expose hillsides, making them more vulnerable to landslides. Natural drainage is impeded
233
managing
urban risk
However, the Indonesian economy
began to break down from mid-1998,
largely as a result of rapid political
change and insecurity. This put large
numbers of people out of work,
pushed up food and other prices and
made it difficult for the government
to obtain foreign credit to finance
food imports. In the capital, Jakarta,
levels of malnutrition increased, and
it became clear that a large
proportion of the population was
suffering from food insecurity.
was put under considerable
pressure. NGO-managed food
distribution was not a realistic
alternative, because of the scale of
the operation and the fact that the
NGOs with relevant experience were
working in rural areas.
14
Indonesia’s food security structures
came under great pressure during
the 1997–98 El Niño event, when
drought reduced rice production and
additional imports were needed. The
government had sufficient
operational capacity and access to
financing to ensure that food was
purchased and distributed, while
donor agencies also embarked on
food assistance programmes in rural
areas, and large-scale severe food
insecurity was largely avoided.

22. Good Practice 4th
8/3/04
2:44 pm
Page 234
©Intermediate Technology/Colin Palmer
disaster risk reduction: mitigation and preparedness
when flood plains are built
upon. Unregulated construction leads to unsafe buildings, and here the middle
classes are also likely to be
vulnerable: many of the fatalities in earthquakes have
been in badly-built apartment blocks.
Urban life has a profound
impact on livelihoods and
coping strategies. Towndwellers rely on cash income
from their labour to a much
Urban vulnerability: a hillside urban slum
greater extent than those
in Lima, Peru
who live in the countryside.
Livelihood strategies are
therefore heavily based on finding paid work. Moreover, the poor are more
likely to undertake dangerous work.
Social capital in the form of community organisations and support networks may
be weak in new, relatively unstable urban communities, especially those that
comprise many different social groups. Newcomers are likely to lack experience
and understanding of specifically urban hazards and avoidance strategies.
Dependence on public support systems and structures (power, water
supplies, public transport infrastructure and emergency services) and
external food supplies is much greater in urban areas. Politicians and officials
have to be lobbied to make improvements to services, but the poor tend to
have the least political influence. City and municipal authorities often lack the
resources and political will to control pollution, provide adequate infrastructure and services, make suitable land available to poor people and implement large-scale mitigation measures.
14.2 Urban risk reduction: constraints and opportunities
14.2.1 Local-level mitigation
The preceding discussion suggests that options for reducing urban risk
through local-level initiatives are limited. It is true that only action by municipal authorities can reduce pollution and other sources of risk on a significant
234

23. Good Practice 4th
8/3/04
2:44 pm
Page 235
chapter 14 managing urban risk
Box 14.1
Physical vulnerability to natural hazards in urban areas
•
•
•
•
•
•
Older residential buildings in
densely-populated areas, which
have been poorly maintained,
altered or extended, or are
overcrowded.
Buildings erected before
adequate standards and controls
were designed, or that have been
built without observing
standards and regulations.
Unplanned, ‘informal’
settlements in marginal, hazardprone areas, where houses do
not incorporate safety features.
Modern buildings of poor design
or construction quality.
Communication and control
centres concentrated in one area.
Hospital facilities that are
insufficient for treating large
numbers of casualties or which
may not be accessible in a
disaster.
Schools and other community
buildings that have been built to
low construction standards or
which cannot be used as
•
•
•
•
•
•
emergency shelters.
Poorly designed or badly built
roads, railways, bridges and
viaducts, embankments and
culverts, whose collapse could
prevent access by the emergency
services, movement of relief
supplies and evacuation of
casualties.
Narrow streets that become
blocked with debris and stalled
traffic, also impeding emergency
assistance.
Water mains that are liable to
rupture, resulting in pollution
and disease.
Sewers that flood, spreading
disease.
Electrical supply lines and
systems, which are liable to
failure.
Gas mains that rupture, with the
risk of fires.
Industrial facilities that are
damaged, leading to leakage of
hazardous chemicals or to fires
and explosions.
Institution of Civil Engineers, Megacities:
Reducing Vulnerability to Natural
Disasters (London: Thomas Telford, 1995),
pp. 17–18.
scale, ensure the resilience of public infrastructure, manage mass evacuations and maintain emergency services, and – with additional support from
national governments – reduce the pressure from the many underlying socio-
235
managing
urban risk
•
•
14
Buildings and physical infrastructure
which may be vulnerable to the effects
of natural hazard events include:

24. Good Practice 4th
8/3/04
2:44 pm
Page 236
disaster risk reduction: mitigation and preparedness
economic forces that make urban dwellers vulnerable. Nevertheless, it is
possible to make a difference at local level, and even relatively small-scale
interventions can have a meaningful impact. The kinds of improvement that
can be made at this level include the following.
Better water and sanitation
Reliable supplies of clean water and adequate sanitation and sewerage
systems are essential to reduce water-related infections. Low-cost technologies for putting up stand-pipes and building latrines and drains have long
been available and, with full community participation, extensive water and
sanitation systems can be installed (see Case Study 14.3).
The importance of making such facilities disaster-resistant is sometimes
overlooked, but this too is relatively simple in technical terms. For example,
in areas regularly visited by floods, water pipes and pumps can be raised
above anticipated water levels, latrines can be sited where they are less likely
to be flooded, drains and sewers can be designed to cope with large volumes
of water, and regularly-spaced manholes can give easy access to drains and
sewers if they need to be unblocked.
Improved storm and surface water drainage
This reduces flooding and the risk of water-borne diseases, and eliminates
breeding sites for mosquitoes and other disease vectors. Blockage of drains
by plastic bags and other rubbish is a major cause of urban flooding, but can
easily be resolved through community action.
Technical improvements should be planned with careful thought to their
consequences. For example, slum dwellers in Indore, India, interviewed in
1994, reported that flood drainage had worsened since the replacement of
open drains with closed ones. The open drains were larger and less easily
blocked, and rubbish and excreta left on the streets were washed into them
to be taken away. Some of the new – blocked – drains ‘backed up’, depositing
water and sewage in people’s homes.7
Management of solid waste
The main element of this is removal of garbage to reduce the risk of disease
and prevent blocked drains. Some activities, such as sorting household
waste, can be managed relatively easily at community level.
236

25. Good Practice 4th
8/3/04
2:44 pm
Page 237
chapter 14 managing urban risk
Case Study 14.3
Community-based sanitation
The Orangi Pilot Project (OPP) in the
city of Karachi, Pakistan, is wellknown internationally as an example
of a large-scale, long-term initiative
based on the skills and resources of
the urban poor. Established in the
1980s in Orangi, one of Karachi’s
poorest districts, the OPP’s
programme of building low-cost
underground sewers had, by 2001,
benefited over 92,000 families
(almost 90% of the settlement) and
the community had invested over
82m rupees (approximately
£924,000) in the system. With the
elimination of the old open sewers,
infant mortality rates fell greatly
(from 130 per 1,000 live births in
1982 to 37 per 1,000 in 1991),
mobility was easier and the cleaner
open spaces in front of houses
created safer play areas for children
and space for social contact and
recreation.
OPP’s achievements have attracted
considerable interest from aid
agencies and led to many attempts
to replicate the approach elsewhere,
with varying degrees of success. OPP
has provided training and advice to
municipal government and other
projects. It has also achieved some
striking successes in lobbying
against proposals for donor-funded
sewerage schemes and presenting
lower-cost alternatives.
A. Zaidi, From the Lane to the City: The
Impact of the Orangi Pilot Project’s Low
Cost Sanitation Model (London:
WaterAid, 2001), www.wateraid.org.uk/
site/in_depth/in_depth_publications.
14
managing
urban risk
Although Orangi’s sloping terrain
helped drainage and hence made
construction of the sewerage
system easier and cheaper than it
would have been on flatter ground,
the main reason for the initiative’s
success has been OPP’s firm
insistence on making community
organisations the primary players
and refusing external funding, which
it believes would make communities
dependent on outsiders and make
projects less cost-effective. Although
OPP provides technical advice and
encouragement, community
organisations direct the projects,
which use local labour and materials.
The communities are responsible for
raising the funds to pay for the work.
237

26. Good Practice 4th
8/3/04
2:44 pm
Page 238
disaster risk reduction: mitigation and preparedness
Flood and landslide defences
A number of appropriate-technology measures can be implemented at local
level. For example, many urban households and communities implement smallscale structural improvements, temporary or permanent, to protect against
flood water and the pollutants it carries: houses are put on raised plinths, or
where this is too costly doorsteps and house fronts are raised; concrete slabs or
rubble and other landfill materials are used to create paths through standing
water; inside houses, there are shelves and raised platforms to store goods
safely, and electricity connections are put at head height.
Flooding can also be curbed by planting trees and other vegetation to absorb
rainwater; the creation of open spaces assists this. Landslides have been controlled by building retaining walls out of old car tyres (see Chapter 10.2.2,
page 147).
Creating safer living environments
This can take many forms. Within the home, use of alternative household fuels
can reduce the risk of indoor air pollution and fires. Better-quality housing provides greater resilience to certain types of hazard. Less crowded settlements
minimise risks to large groups of people and provide escape routes and safe
places, but this can be difficult to manage where pressure on land is great.
Relocation, and avoiding hazardous areas
These measures are often suggested as a solution to disaster risk, but there
are serious problems associated with them. Relocation to less hazardous
areas is highly effective in reducing physical risks, but overlooks the centrality of livelihood security to vulnerable people and the intensity of the socioeconomic pressures that drive them to occupy hazardous land.
People settle in hazardous locations for a variety of reasons, but the main
ones are livelihood-related:
• rents are cheaper;
• they are closer to places of employment and job opportunities;
• there is access to markets (for buying and selling), health services,
schools, electricity and water; and
• in central locations, they are more visible to the public, which means that
there is a better chance of receiving funds for improvements.
238

27. Good Practice 4th
8/3/04
2:44 pm
Page 239
chapter 14 managing urban risk
Relocation schemes, on the other hand, tend to involve sites on the edges of
cities. Land may be more readily available and affordable here, but jobs and
many other facilities are not. For this reason, relocation is often resisted by
vulnerable groups, although pressure can be brought to bear on them to
move. In some cases, there are suspicions that relocation of communities is a
pretext for land to be cleared for more commercial development.8 Handled
sensitively, and with full community participation, relocation can bring benefits, but it is a complex process (see Case Study 14.4).
Better working environments
Improved health and safety practices reduce the risk of disease and injury
from unsafe working conditions and exposure to hazardous materials.
Lobbying for better environmental management
This includes calling for improved water and sanitation, health and safety at
work, and curbs on polluting industries and practices. It means targeting private-sector polluters as well as public authorities. This is a necessary task but
a difficult one, as the private sector is powerful and resistant to pressure,
while public authorities may be unable or unwilling to support community
groups to challenge industrial practices that create risk.
There are instances of community organisations fighting successfully against
pollution and commercial developments that threaten them, but there are
many more examples of failure. For example, one-third of industrial enterprises surveyed in the city of Lahore, Pakistan, had received complaints about
pollution from local groups and people, but none had made any significant
efforts to improve their practices.9
Urban agriculture
14
They are also well placed to exploit the potential of urban agriculture, which
is surprisingly extensive, especially in Africa and on the edges of towns and
cities, and has grown significantly since the late 1970s in response to need.
Globally, about 200m urban dwellers are also urban farmers, providing food
and income to an estimated 700m people. Evidence from cities across the
managing
urban risk
Towns and cities largely depend on actions at national and even international
levels to ensure adequate and affordable supplies of food; NGOs do not have the
capacity to manage acute, large-scale urban food crises (see Case Study 14.2).
However, local-level agencies can play a role in food monitoring and distribution.
239

28. Good Practice 4th
8/3/04
2:44 pm
Page 240
disaster risk reduction: mitigation and preparedness
Case Study 14.4
Community participation in relocation
The city of Santo André, with 62,500
inhabitants, is part of the
metropolitan region of São Paulo,
one of Brazil’s largest industrial
centres. About 15% of the population
lives in slums located in flood- and
landslide-prone areas.
In 1997, under its slum upgrading
programme, the municipality
identified the Sacadura Cabral
settlement, where 780 families lived,
as a priority area. The settlement
was prone to flooding every year and
was very densely populated (853
people/hectare). To achieve
upgrading and reduce flood risk, it
was necessary to level the area with
earth and relocate some 200 families
to nearby areas.
The initial top-down, random-selection
method for identifying families to be
moved was strongly resisted by local
people, so it was dropped and
replaced with a series of meetings
with communities and their leaders. A
new proposal was then developed
jointly, which comprised:
•
construction of 200 apartments
in a housing complex one
•
•
•
kilometre from Sacadura Cabral,
and the relocation of 200
families to live in them (at the
end of 1998);
subsidised credit for the families
that acquired the apartments;
demolition of the relocated
families’ homes and levelling of
the land, followed by a process
of internal relocation within the
settlement and levelling of the
rest of the flood-prone area;
reconstruction of homes by the
people concerned, with technical
assistance from the local
government.
Selection of the families to be
moved was undertaken in
consultation with the community, on
the basis of agreed criteria
including income levels and family
size. Several planning meetings
were held throughout the
settlement.
M. G. G. Oliveira and R. Denaldi,
‘Community Participation in Relocation
Programmes: The Case of the Slum
Sacadura Cabral in Santo André – Brazil’,
Open House International, vol. 24, no. 3,
1999, pp. 24–32.
South has shown that urban agriculture can have a considerable impact on
local food production and consumption, as well as livelihoods. There is
240

29. Good Practice 4th
8/3/04
2:44 pm
Page 241
chapter 14 managing urban risk
probably much that can be done to promote and develop it. At the same time,
urban agriculture can bring its own hazards, such as creating breeding sites
for mosquitoes through irrigation, exposure to pesticides, contamination of
crops from the presence of lead and other heavy metals in the soil, and
contamination from the use of human excrement as fertiliser.10 The potential
gains and risks of urban agriculture projects should be considered carefully
by organisations working at local level.
14.2.2 Community-based approaches
Formal and informal social organisations exist in every part of a town or city
where people live and work. But the notion of a ‘community’ may be quite
different from rural areas, because in an urban district there is often a
complex mix of very different social groups within a small area. The locus of
organisation for people may be their neighbourhood – the district where they
live – rather than a social community.11 This presents problems to NGOs and
other agencies trying to establish locally-based mitigation initiatives. The
mutual suspicion that often exists between officials and people living in
informal settlements adds to the difficulty.
It is useful to distinguish between ‘private’ and ‘public’ space when implementing community-level risk reduction measures in urban areas. Most
241
14
Participatory initiatives (see Chapters 7.4 and 8) can help to break down
barriers between social groups as well as identifying opportunities for locallevel action. Community vulnerability analysis and action plans can be used
in advocacy, and other stakeholders – city government departments, NGOs,
academic institutions – can be brought into urban action planning
processes.13 There are indications that such approaches have galvanised
community action in risk reduction, and stimulated greater cooperation with
officials (see Case Study 14.5), but more thorough documentation and
analysis are needed to understand how this process works in different
contexts.
managing
urban risk
Nevertheless, many communities are willing to undertake risk-reducing
measures collectively where they are able to do so. For example, a survey of
households in the Albouystown district of Georgetown, Guyana, found that,
even though relatively few people were members of community-based organisations, 63% of those interviewed cleaned drains with their neighbours, and
53% cleaned streets with their neighbours.12 The Orangi Pilot Project (Case
Study 14.3) and other schemes inspired by it in Pakistan demonstrate that
community-based action is possible on a relatively large scale.

30. Good Practice 4th
8/3/04
2:44 pm
Page 242
disaster risk reduction: mitigation and preparedness
Case Study 14.5
Building bridges in Lima
The Caqueta district in Peru’s capital,
Lima, grew rapidly in the late 1980s,
and by 1999 was estimated to be
housing 15,500 people living in
buildings ranging from wooden
shacks to constructions of adobe
and concrete. A 30-foot-deep ravine
formed by the Rimac river crosses
the district. Landslides at its edge
are frequent, damaging and
destroying squatter housing.
Several initiatives have been
discussed to solve the problem of the
ravine. They include covering it with a
lightweight structure on which
industrial enterprises could be built,
and relocating residents and removing
all buildings and infrastructure. But
none of these essentially large-scale,
technological approaches has been
implemented, owing to their
complexity and cost, lack of political
will and administrative coordination,
and failure to consult with local
people, who do not want to move
away from their places of work.
A different approach was attempted
by a local NGO, Instituto Para la
Democracia Local (IPADEL), with
support from external consultants.
The project team carried out an
evaluation of local hazards,
vulnerabilities (physical and social)
242
and capacities, based on surveys,
questionnaires and participatory
workshops with market and housing
association representatives.
The evaluation’s findings fed into a
three-day planning workshop
attended by representatives of
municipalities, traders’ and residents’
associations, local NGOs and the fire
services. Several ideas for reducing
risk were put forward. Two were
accomplished by the project. The first
was to improve the pedestrian bridge
across the ravine by installing
protective sides high enough to
prevent children from falling off.
The second initiative was to reduce
fire risk in the local markets. Several
planning and training workshops were
held involving traders and firemen.
One consequence of the work was an
improvement in relations between the
two groups. Fire-fighters had
previously been seen as part of the
municipal authority and, given a
tradition of tensions between the
authority and market traders, as
hostile to the traders.
D. Sanderson, ‘Implementing Action
Planning to Reduce Urban Risk’, Open
House International, vol. 24, no. 3, 1999,
pp. 33–39.

31. Good Practice 4th
8/3/04
2:44 pm
Page 243
chapter 14 managing urban risk
people will take steps to protect their own homes, as far as they are able, but
the extent of their responsibility for protecting public space may be less clear.
Urban residents are likely to feel that it is the municipal authorities’ responsibility to take care of streets, drains and other communal facilities. Yet the
boundary between private and public space may be drawn differently in
different areas. In one community, keeping lanes clear of refuse and debris
may be seen as a community responsibility; in another, it may be seen as the
authorities’ task. A community’s willingness to take on responsibility for
managing environmental risks also depends on how far it feels government
should, and can, do so.
14.2.3 Partnership
Local-level activity must be linked to town- or city-wide measures to improve
services and mitigate the impact of hazards. Without this, local initiatives will
always be threatened by external pressures. Towns and cities are affected by
pressures coming from beyond their boundaries, but local organisations can
only address these through advocacy. This makes it particularly important to
develop risk management partnerships between communities, grass-roots
organisations, NGOs, municipal authorities and others (see Case Study 14.6).
The introduction of Local Agenda 21s – programmes for addressing environmental problems, mandated by the Agenda 21 programme agreed at the
Earth Summit in Rio in 1992 – has created space for better partnerships by
moving environmental planning into the public arena, providing for more
extensive public consultation and partnerships. Local authorities are being
encouraged to think less about doing everything themselves, and instead to
put greater effort into creating ‘enabling frameworks’ that facilitate partnerships between groups. Within its comprehensive environmental coverage,
Agenda 21 includes problems of human settlement planning and management in disaster-prone areas.14
243
14
Much of the attention being given to disaster risk in urban areas has concentrated on the problems of cities. Here, the scale of the problems makes
involvement of municipal and even central governments particularly important. But most of the urban population in developing countries does not live
in cities but in smaller market towns and administrative centres. It is in such
places that most urban growth is taking place. Here, it can be easier to create
the partnerships needed to tackle problems at local level – and it is arguably
managing
urban risk
A number of other international initiatives address aspects of urban risk
reduction, focusing particularly on municipal authorities (see Box 14.2).

32. Good Practice 4th
8/3/04
2:44 pm
Page 244
disaster risk reduction: mitigation and preparedness
Case Study 14.6
Collaborative action planning for earthquakes
The Kathmandu Valley Earthquake
Risk Management Project (KVERMP)
is one of a number of initiatives in
South and South-East Asia under the
Asian Urban Disaster Mitigation
Programme (see Box 14.2). The
Kathmandu Valley, with a population
of over a million people, is
vulnerable to earthquakes. It has not
suffered a major earthquake since
1934, but many experts believe that
another is long overdue.
KVERMP has addressed several
aspects of risk reduction by
developing a scenario and action plan,
making schools safer by retrofitting,
raising public awareness and building
the capacity of local institutions.
The first step was to assess the
valley’s earthquake risk. The project
did not attempt to develop a new,
improved assessment of this risk,
which would have been a complex,
costly exercise. Instead, it chose to
present what was already known
about that risk in a form that was
understandable to officials and the
public. This made it easier to involve
a range of groups in the subsequent
action planning. About 80 different
organisations, mostly from
government but also including NGOs,
academic institutions, businesses
and donors, were involved in
244
creating the action plan. Many
different risk reduction measures
were proposed, and project staff
visited participating organisations to
discuss these, and the feasibility of
including them in the plan. A draft
plan was then distributed to all the
stakeholders for comment.
The final draft was presented at a
two-day workshop attended by 85
representatives of all the
organisations involved, who
discussed its approach and contents
and helped to set priorities. The
agreed action plan was printed and
launched by the Prime Minister at an
event to mark the country’s annual
Earthquake Safety Day, and
distributed widely. It is used
regularly in training and orientation
programmes, and has generated
considerable debate and feedback
among local inhabitants, policymakers, administrators and scientific
and technical specialists.
A. M. Dixit et al., ‘Mitigation Planning and
Implementation: Experiences of KVERMP’,
in Proceedings: Regional Workshop on
Best Practices in Disaster Mitigation.
Lessons learned from the Asian Urban
Disaster Mitigation Program and Other
Initiatives, 24–26 September 2002, Bali,
Indonesia (Bangkok: Asian Disaster
Preparedness Center, 2002), pp. 181–87.

33. Good Practice 4th
8/3/04
2:44 pm
Page 245
chapter 14 managing urban risk
Box 14.2
International initiatives in urban risk reduction
245
In 1998–99, the UN ISDR’s Risk
Assessment Tools for Diagnosis of
Urban Areas against Seismic
Disasters (RADIUS) project worked in
nine cities to develop risk
assessment methods, raise public
awareness and produce action plans.
Post-project evaluations have
provided examples of the action
plans being used for emergency
training, urban development
planning and updating official
disaster management structures. The
assessment tools are also being
adopted by other towns and cities.17
Local Authorities Confronting
Disasters and Emergencies (LACDE)
is a worldwide network of local
authorities and other organisations
seeking to increase local
government’s understanding and
capacity to deal with disasters of all
kinds. Its main activities are wellattended annual conferences and a
newsletter.18
managing
urban risk
The Asian Urban Disaster Mitigation
Programme (AUDMP), managed by
the Asian Disaster Preparedness
Center in Bangkok, is a six-year
initiative covering eight countries in
South and South-East Asia. It aims to
establish sustainable public and
private sector mechanisms for
disaster mitigation, and to promote
the replication and adaptation of
successful approaches. Its activities
comprise demonstration projects in
each country to assess, design and
implement measures to mitigate a
number of natural and technological
hazards, information-sharing and
networking, and training.16
14
The Healthy Cities Programme, run by
the World Health Organisation (WHO),
attempts to improve public health
through better control of
environmental and other hazards. It
seeks to put health higher on
decision-makers’ agendas, build up a
lobby for better public health and
develop popular, participatory
approaches to dealing with health
problems. At town/city level, it is
implemented by local teams drawn
from government, NGOs, communities
and other organisations, with
technical back-up from WHO offices.
Several hundred towns and cities have
received the ‘Healthy City’ designation
– i.e. they are undertaking or planning
new initiatives involving collaboration
between government and community
organisations. These have included
upgrading poor people’s housing,
improving sanitation in informal
settlements, improving drainage to
prevent flooding, better waste
management, and more integrated
planning.15

34. Good Practice 4th
8/3/04
2:44 pm
Page 246
disaster risk reduction: mitigation and preparedness
more vital to do so, in view of the relative neglect of this issue in such towns
and the often limited capacity and resources of local government for
managing disaster risk.
Environmental deprivation is increasingly being recognised as an important
dimension of urban poverty, which means that poverty reduction
programmes are considering environmental issues, including hazards. Local
Agenda 21s have furthered this process. Opportunities to bring the poverty,
environmental and risk management agendas closer together are probably
better here than in other contexts.
14.3 Chapter summary
• Rapid and uncontrolled urbanisation is putting ever-greater numbers of
people at risk from natural and technological hazards.
• Environmental hazards are the main cause of ill-health, injury and premature death in urban areas in developing countries and there is a significant risk of major urban disasters in future.
• Man-made and technological hazards are particularly important in towns
and cities.
• Poor and marginalised people are more likely to live in unsafe areas and
housing, and in densely populated districts where the impact of hazards
and disease is magnified.
• Urban vulnerability is distinctive in terms of the composition of livelihood
and coping strategies, the nature of social capital and the level of dependence on public support systems and structures.
• Only action by municipal authorities can reduce risk on a significant scale,
but it is possible to make a difference at local level in, for example,
improving water, sanitation and drainage systems, constructing flood and
landslide defences and creating safer living environments.
• Avoidance of hazardous areas and relocation to safer ones are often
suggested as a solution, but are rarely feasible given the intensity of the
socio-economic pressures that force people to live in dangerous locations.
• There are many opportunities for partnerships between communities and
other groups.
Notes
1
IIED, Briefing Paper Series on Urban Environmental Improvement and Poverty Reduction
(London: International Institute for Environment and Development, 2001),
www.iied.org/rural_urban/moredownloads.html#ueipr.
246

35. Good Practice 4th
8/3/04
2:44 pm
Page 247
chapter 14 managing urban risk
14
managing
urban risk
2 A. J. McMichael, ‘The Urban Environment and Health in a World of Increasing Globalization:
Issues for Developing Countries’, Bulletin of the World Health Organization, vol. 78, no. 9,
2000, p. 1,121.
3 Cities of the Future: Dream or Nightmare? (London: Panos (Briefing 34), 1999), www.panos.org.uk.
4 Ibid.
5 C. Kelly, Acute Food Insecurity in Mega-Cities: Issues and Assistance Options (London: Benfield
Hazard Research Centre (Working Paper 7), 2003), www.benfieldhrc.org.
6 Cities of the Future.
7 C. Stephens, R. Patnaik and S. Lewin, ‘This Is My Beautiful Home’: Risk Perceptions Towards
Flooding and Environment in Low-Income Urban Communities. A Case Study in Indore, India
(London: London School of Hygiene and Tropical Medicine, 1995), pp. 36–37.
8 See for example J. Twigg (ed.), Living with Disaster (Rugby: ITDG), pp. 11–14; World Disasters
Report 2002: Focus on Reducing Risk (Geneva: International Federation of Red Cross and Red
Crescent Societies, 2002), pp. 54–55.
9 R. Hameed and J. Raemaekers, ‘Putting Over the Message: A Programme in Pakistan To Build
Capacity among Industrialists for Pollution Abatement’, in C. Gandelsonas (ed.),
Communicating for Development: Experience from the Urban Environment (London: ITDG
Publishing, 2002), pp. 187–97.
10 L. Mougeot, ‘Overview – Urban Food Self-Reliance: Significance and Prospects’, IDRC Reports,
vol. 21, no. 3, 1997, www.idrc.ca/books/reports/V213/overview.html; A. J. McMichael, ‘The
Urban Environment and Health in a World of Increasing Globalization: Issues for Developing
Countries’, Bulletin of the World Health Organization, vol. 78, no. 9, 2000, p. 1121.
11 B. Wisner, ‘Social Planning and Activism for Earthquake Risk Reduction’, in C. Rodrigue and E.
Rovai (eds), Earthquakes (London: Routledge, in press).
12 M. Pelling, The Vulnerability of Cities: Natural Disasters and Social Resilience (London:
Earthscan, 2003), pp. 131–37.
13 M. Bhatt, M. Gupta, A. Sharma, ‘Action Planning – From Theory to Practice’, Open House
International, vol. 24, no. 3, 1999, pp. 16-23.
14 Agenda 21, UN Division for Sustainable Development,
www.un.org/esa/sustdev/agenda21text.htm.
15 WHO Healthy Cities: A Programme Framework (Geneva: World Health Organisation, 1995),
www.who.int/hpr/archive/cities/docs/hc_framework.pdf; Building a Healthy City: A
Practitioners’ Guide (Geneva: WHO, 1995),
http://www.who.int/hpr/archive/cities/docs/hc_guide.pdf.
16 The AUDMP’s website is: www.adpc.ait.ac.th/audmp/aboutaudmp.html.
17 RADIUS’ website is www.unisdr.org/unisdr/radiusindex.htm.
18 LACDE’s website is: www.ulai.org.il/lacde.htm.
247

36. Good Practice 4th
8/3/04
2:44 pm
Page 248
Chapter 15
Slow-onset disasters
15.1 Introduction
Writing on disasters usually distinguishes between rapid-onset events and
slow-onset events. The critical factor here is said to be the nature of the
hazard concerned.
Hazards that arise suddenly, or whose occurrence cannot be predicted far in
advance, trigger rapid-onset disasters. Earthquakes, cyclones and other
windstorms, landslides and avalanches, wildfires, floods and volcanic eruptions are usually categorised as rapid-onset events. The warning time ranges
from seconds or at best a few minutes in the case of earthquakes and many
landslides, to several days in the case of most storms and floods. Some
volcanic eruptions may be preceded by weeks or months of activity, but
predicting volcanoes’ behaviour remains very difficult and the warning time
for the eruption itself may be only days or hours. Most disasters are rapidonset events.
Most discussion of slow-onset disasters concentrates on one hazard: drought.
It can take months or sometimes years for the results of drought to become
disastrous, in the form of severe water and food shortages and, ultimately,
famine. Drought is not the only relevant hazard, though. Pollution of the environment can also be considered a slow-onset disaster, particularly in cases of
growing concentrations of toxic wastes, which may build up over years. Human
activities that degrade the environment and damage ecosystems – deforestation for instance – also contribute to disasters. Their cumulative impact may not
be felt for decades, although the hazards that they make more likely, such as
flash-floods and landslides, may be sudden-onset events.
To some extent, the distinction between slow- and rapid-onset disasters is
artificial. Hazards certainly can be categorised in this way. Disasters, on the
other hand, are the product of hazards and human vulnerability to them. The
socio-economic forces that make people vulnerable may act quickly or slowly,
but in most disasters it is likely that long-term trends will be more influential.
When viewed in this light, it could be argued that all disasters are slow-onset.
This Good Practice Review argues for a long-term, holistic approach towards
managing risk that also breaks down the distinction between slow- and rapidonset disasters.
248

37. 2:44 pm
Page 249
chapter 15 slow-onset disasters
Why then have a separate chapter on slow-onset disasters? There are two
related reasons. First, because real-life disaster management often does
distinguish between rapid- and slow-onset events. The approaches that have
been developed for dealing with slow-onset hazards such as drought may
differ from those used against other hazards. The documentation of such
work is generally much better here than in other disasters, and there is a
greater volume of high-quality research and analysis.
The second reason is that, in the management of drought and food security,
and in environmental renewal initiatives, development agencies play a much
more significant role than they do in sudden-onset disasters. This may offer
lessons about how to get the development community more involved in
disaster management generally. Mitigation of drought and food security is
more advanced than mitigation of many other types of disaster in integrating
livelihood issues with disaster management, adopting participatory and
community-level approaches, and building upon indigenous knowledge,
coping strategies and appropriate technologies.
15.2 Environmental degradation and pollution
15.2.1 Environmental degradation
There is a strong link between environmental degradation and increased risk
from natural hazards. Bad management of natural resources and destruction of
the eco-system make disasters more likely. For example, the removal of trees,
bushes and other vegetation in the course of building, farming or other
commercial activities can create hazardous conditions. In agricultural areas it
accelerates the loss of fertile topsoil to wind and water erosion. Water is no
longer held in the soil by vegetation and so runs away rapidly, which increases
vulnerability to drought. On hillsides, rapid water run-off can cause flash-floods
and landslides. In coastal zones, the destruction of mangrove forests removes a
natural barrier to the winds and sea surges created by tropical cyclones.
In all these cases, environmental protection or renewal is technically feasible.
Natural resources can be managed and replenished through such measures as
reforestation and other planting, waste management, environmentally sustainable farming and grazing practices, terracing, and building protective stone and
earthworks to prevent rapid water run-off. Countless projects of this kind have
been undertaken. Complementary activities include environmental education.
Attempts to protect the environment often challenge powerful interests that
stand to gain from its destruction: timber companies from logging; ranchers
249
slow-onset
disasters
8/3/04
15
Good Practice 4th

38. Good Practice 4th
8/3/04
2:44 pm
Page 250
disaster risk reduction: mitigation and preparedness
from stripping land of woods, hedges and bushes; shrimp farms and hotels
from tearing up mangroves. The example of the Afro-Honduran Garifuna
communities in Chapter 12 illustrates how difficult and even dangerous it can
be to make such challenges. More generally, economic and demographic
pressures on poor countries, coupled with entrenched political and cultural
attitudes, inhibit an effective response to recognised environmental crises.
The failure of the Soviet Union and the post-Soviet Central Asian republics to
deal with the drying up of the Aral Sea is a good example of this.1
15.2.2 Pollution
Major industrial disasters are often rapid-onset: the result of industrial or
transport accidents causing explosions and fires of flammable material or the
release of oil, chemicals or radioactive material. The chemical leak at Bhopal
in India in 1984 and the accident at the Chernobyl reactor in Ukraine in 1986
are the best known recent examples. Although the number of such industrial
disasters remains low and there are still relatively few casualties compared to
those from natural hazards, their numbers are rising with the spread of industrialisation in developing countries.2 However, the impact of pollution may
not be felt for decades as volumes of waste accumulate and growing
numbers of people are exposed.
Although local-level management of some forms of waste and pollution is
feasible, significant reductions in pollution and improvements in waste
management require the involvement of the state in setting and enforcing
standards and providing adequate public facilities. Communities and their
organisations can encourage this through vigorous advocacy (see Chapter
12). In many instances, specialist scientific and engineering expertise will be
required, especially in the case of extremely toxic wastes.
Community organisations, NGOs, researchers, governments and other
agencies should work together to tackle major problems of this kind collectively – the ‘partnership’ approach to disaster management set out in Chapter
5. The difficulty of creating successful partnerships should not be underestimated, especially in the case of industrial pollution where the demand for
better protection may conflict with business aims. Case Study 15.1 shows
how agencies have responded to a major crisis resulting from groundwater
pollution.
The connections between climate change and hydro-meteorological hazards
such as droughts and cyclones are becoming apparent, backed up by a
growing body of scientific evidence.3 Although climate change specialists and
250

39. 8/3/04
2:44 pm
Page 251
15
chapter 15 slow-onset disasters
Case Study 15.1
Arsenic pollution in Bangladesh
In the early 1990s, high
concentrations of arsenic were
reported in groundwater in western
Bangladesh. Arsenic is a cumulative
and potentially fatal poison, and
many of the more advanced
symptoms of poisoning are
incurable. About 25m people have
been exposed through drinking
water from tube wells, and over
7,000 cases of arsenic poisoning
have been confirmed. Fifty-nine of
the country’s 64 districts have some
arsenic-contaminated groundwater.
There has been considerable debate
about the source and release
mechanisms. It is now widely
accepted that the arsenic is of
natural, geological, origin, although
it is less certain how it gets from
sediments into groundwater.
In response to a potentially massive
disaster, international agencies, the
government of Bangladesh, scientists
and local NGOs have put considerable
effort into researching the problem
and identifying and implementing
solutions. The costs of this work run
into millions of dollars. Activities
undertaken to date have included:
•
•
•
•
•
•
•
testing water supplies (using
specially developed field test
kits) and monitoring cases of
arsenic poisoning;
data collection and analysis
based on geochemical surveys
and further research on the
mechanisms by which arsenic is
released;
promoting methods of removing
arsenic from water (e.g. filters)
and researching new methods;
finding arsenic-free sources of
water (e.g. digging deep tube
wells that go beneath the
arsenic-bearing sediments,
rainwater harvesting);
training doctors and health
workers to identify and treat
arsenic poisoning, and
developing patient treatment
protocols;
public education; and
information dissemination and
networking among professionals.
E. Jones, Arsenic 2000: An Overview of
the Arsenic Issue in Bangladesh (Dhaka:
WaterAid Bangladesh, 2000),
www.wateraid.org/site/in_depth/
research_centre/groundwater/183.asp.
disaster managers have begun talking to each other, the way forward in
coping with the threat remains far from clear.
251
slow-onset
disasters
Good Practice 4th

40. Good Practice 4th
8/3/04
2:44 pm
Page 252
disaster risk reduction: mitigation and preparedness
At a global level, advocacy to reduce greenhouse gas emissions is essential,
and there is already a substantial alliance of scientists, environmentalists
and some businesses (notably insurers) engaged in this. At local levels,
disaster managers are unsure what they can do about climate change beyond
what they are already doing to minimise risk. How can they calculate the
increased risk due to this problem? How far should their existing disaster
planning be stepped up to counter the threat from climate change? As yet,
there are no answers to these questions, but they are needed urgently.
15.3 Drought, food security and famine
©ITDG/Annie Bungeroth
More than 70m people died in famines during the twentieth century.4 There
have been many major, high-profile food crises over the past 30 years. Most
have been in Africa. Many have been triggered by drought, but other hazards
including floods, harsh winter weather and diseases that affect crops and
livestock can also act as triggers.
Sudan: harvesting aubergines grown in holes
to collect water, reduce evaporation and
provide shade
Natural hazards are only one factor in food
crises. Political, economic and social factors,
including conflict, can be powerful contributors. The macro-economic policies of
governments and international agencies
have played a considerable role in creating
food insecurity and famine (see Case Study
2.2, page 18). In some countries, food aid
has become a regular, even continuous,
component of government development
plans and programmes: Ethiopia and Iraq
are prominent examples, though for
different reasons. The food crisis in
Southern Africa in 2002 revealed the influence of the HIV/AIDS pandemic on food
insecurity. Research on the likely consequences of global warming indicates that
areas already prone to drought are likely to
suffer even more severely in future.
These are massive challenges to overcoming famine and food insecurity. Many
argue that the challenges are growing and
becoming more complex. However, understanding of the causes of food insecurity
252

41. 2:44 pm
Page 253
chapter 15 slow-onset disasters
and famine has advanced greatly in the past 30 years or so. Also during that
time, numerous advances in practice, from local to international levels,
have greatly enhanced the capacity to improve food security and predict
and prepare for crises. The adoption of community-based approaches,
appropriate technologies and indigenous knowledge is significant, as is the
growing integration of food security initiatives with those seeking to
support livelihoods more broadly.
Food security is complex, and the literature on the subject is extensive.
Experience is widespread among development agencies. There is plenty of
good general guidance on issues and operations for project planners, on
which much of this overview is based.5
15.3.1 Understanding ‘food security’
There is no fixed yardstick for measuring food security. It is not simply a case
of people having enough to eat; rather, it is:
when all people, at all times, have physical and economic access to sufficient, safe and nutritious food to meet their dietary needs and food preferences for an active and healthy life.6
This depends on people being able to buy food or obtain it in other ways,
such as in exchange for their services (e.g. labour) or borrowing from
members of their extended family or community. This, in turn, depends on
them having sufficient income, savings and other material assets, skills or
social connections to obtain food. It also depends on external factors such as
the price of food in the market.
Control of the supply and distribution of food is another important dimension.
Some members of a household have little influence over how food is distributed. Children depend on the food they are given by adults. Male children and
adults often get more food than females. Providing adequate food for elderly
family members may not be a priority when times are hard. Sometimes decisions about food distribution are made on the basis of economic rationality –
for instance, a family member who brings in a regular wage may get priority –
but power relationships and cultural practice are also influential.
There can be pockets of food insecurity almost anywhere – within countries,
communities and families. Hence, monitoring and analysis of food insecurity
should take place at different geographical and social levels, as well as at
different times of the year (see Section 15.4, below).
253
slow-onset
disasters
8/3/04
15
Good Practice 4th

42. Good Practice 4th
8/3/04
2:44 pm
Page 254
disaster risk reduction: mitigation and preparedness
Food security is not a mere question of there being enough food available –
rather, it reflects the fact that people do not have equal access to food
because of differences in the resources they possess and other economic,
social and political factors. This theory of people’s differing ‘entitlements’ to
food, first advanced by the economist Amartya Sen in the 1970s, has transformed the way many aid and development agencies view food insecurity and
famine, and has two important implications for any agency seeking to reduce
the risk of food insecurity.
First, it shows that there are many different ways of overcoming food insecurity
by improving people’s ability to obtain food. These include conventional food
production practices (e.g. promoting improved farming techniques), standard
drought mitigation practices (e.g. soil and water conservation), and better
management of natural resources such as forests and watersheds. They can also
include measures to support livelihoods generally, such as projects to create
jobs and increase incomes, savings and credit programmes, improving the
quality of water supplies and sanitation (poor health is an important contributor
to malnutrition), better education for women (known to be an important factor in
reducing malnutrition at household level), helping local markets for food and
other products (through better access roads and footpaths, or better methods of
packaging and preserving perishable products for sale), and encouraging wider
community participation in economic and social development initiatives to
improve the situation of marginalised groups. This opens the way for development agencies to combat food insecurity – and many are doing so.
The second point is that such initiatives can be undertaken locally, be it to
tackle pockets of food insecurity or to contribute to more widespread
programmes. There is ample opportunity for local-level organisations to
become involved.
15.3.2 Drought and food security
Although the causes of food insecurity include political, social and economic
factors, natural hazards, especially drought, remain important.7
Scientists and geographers distinguish between three kinds of drought:
1. Meteorological drought is when rainfall drops below a certain level.
2. Hydrological drought involves a reduction in water resources such as
rivers, lakes and underground water: this too is the result of lower rainfall,
but it may take some time to be felt, and in the case of river systems may
be the result of a drop in rainfall far away from the area in question.
254

43. 2:44 pm
Page 255
©ITDG/Orlando Arnold
chapter 15 slow-onset disasters
3. Agricultural drought is the impact
of the other two kinds on crop yields.
Whilst it is obvious that places that
receive little rainfall are droughtprone, drought can also occur where
rain is normally sufficient – for
example, parts of Bangladesh have
suffered a number of droughts, some
quite severe. A few weeks with little
or no rain at a critical time of the year
for crop growth can be devastating,
even if the rest of the year is not dry.
The important point is whether the
amount of rainfall is sufficient for
agriculture, livestock and other
human needs at the time in question.
Drought’s impact is felt on different
geographical scales. Climatic conditions can vary widely within a
Constructing a traditional hand-dug
country. At local level, changes in
reservoir (hafir) in Sudan
land use can have a significant
impact on watersheds, and hence
the amount of water available through run-off or from rivers and ponds.
Some of the measures needed to mitigate water shortages and their effects are
outlined below. Here, a more general problem should be noted: the difficulty of
assessing the relative importance of drought on food insecurity compared to
socio-economic factors. It can be hard to disentangle these causes, because
they interact with each other. For example, a fall in crop yields may be due to
lack of water (i.e. agricultural drought), but may also be the result of such
factors as a lack of fertilisers or weeding, pests and crop diseases, labour shortages at critical periods and low prices for crops in the markets.
Drought can trigger other hazards: in Ethiopia in 1984, it contributed to an
infestation by army worm that greatly increased crop damage.8
15.3.3 Coping with food insecurity and famine
Seasonal food insecurity is normal in many poor households, which suffer
from a hungry period shortly before the harvest, as food gathered from the
255
slow-onset
disasters
8/3/04
15
Good Practice 4th

44. Good Practice 4th
8/3/04
2:44 pm
Page 256
disaster risk reduction: mitigation and preparedness
previous harvest runs out. This problem is known as ‘chronic’ food insecurity,
as opposed to one-off food shortages, which are classified as ‘transitory’
food insecurity.
When food insecurity is acute and prolonged, it can lead to starvation and
finally to famine (famine is where there is a significant increase in sickness and
death rates resulting from starvation and associated factors). The descent into
acute food shortage and thence into famine can take weeks, months or even
years. Except where conflict is a major contributing factor, famine should not be
seen as inevitable because of the range of measures available at all levels to
improve food security and strengthen livelihoods in the long term.
Take, for example, the mitigation strategies used by communities to protect
themselves against drought and the food insecurity that results from it. These
can be divided into two main types: agricultural and non-agricultural.
• Agricultural mitigation strategies are measures to maintain crop and livestock production. They include sowing again after a crop has been ruined by
drought, sowing alternative crops, or moving livestock to other locations.
• Non-agricultural mitigation strategies include seeking off-farm employment in the locality or elsewhere, eating seeds or roots that were saved to
be sown in the next growing season, reducing the amount of food
consumed, eating wild food such as berries and roots, postponing social
functions such as weddings, using up savings and selling assets (such as
livestock, household goods and personal possessions), buying on credit,
borrowing money or calling in favours from communities and kin.
Families only sell their livelihood assets when they have to, which is when
other methods such as growing alternative crops or finding alternative
employment are insufficient. Sale of assets is a good indicator of how severe
the consequences of drought are. Poor families, those with small landholdings and the landless, are the first to resort to such methods. Only when all
else has failed will whole families and communities migrate in search of food
(e.g. Case Study 9.2, page 137).
The severity of a food crisis can therefore be judged by looking at food and
livelihood coping strategies as well as at food supplies. Food insecurity
among a particular population is likely to be acute if:
• People experience a large reduction in their major source of food and are
unable to make up the difference through new strategies.
• The prevalence of malnutrition is abnormally high for the time of year, and
this cannot be accounted for by health or care factors.
256

45. 2:44 pm
Page 257
chapter 15 slow-onset disasters
• A large proportion of the group is using marginal or unsustainable coping
strategies.
• People are using coping strategies that are damaging their livelihoods in
the longer term or incur some other unacceptable cost such as acting illegally or immorally – stealing, for instance.9
The most effective way to protect communities against food insecurity and
famine resulting from drought is to strengthen these diverse mitigation
strategies well in advance, especially those that enable them to preserve
their productive assets, such as animals, seeds and tools.
Despite this, most external intervention is still in response to drought, not in
helping to create more drought-resistant communities. Moreover, it often
comes at a late stage, when communities are in crisis and may already be
destitute, having been forced to dispose of productive assets. Typical interventions in such circumstances are to provide food, seeds, fertiliser, animals
and agricultural equipment to replace that which has been used or sold, and
to lend money. Where a crisis has become acute, with widespread starvation
and migration, aid agencies’ interventions focus on emergency response,
especially feeding and health care. Food-for-work and cash-for-work schemes
are also common responses.
The boundary between disaster preparedness and response is blurred in food
crises, because they can develop over such a long period. Some would argue
that an emergency begins when hungry people are forced to dispose of their
livelihood assets; others put it at the point where destitute, starving people
leave their homes to beg for food, or even where large numbers begin to die
of starvation. In this chapter, emphasis is placed on longer-term mitigation
measures to maintain food production and incomes.
15.3.4 An integrated approach
The ideal approach to drought-related food insecurity addresses the different
dimensions of the problem, using a range of methods.
An example of this range is given in Table 15.1, which highlights the main
interventions in a joint UNICEF/WHO project that covered 600 villages in
Iringa, Tanzania, in the 1980s. The project was a response to persistent food
insecurity and malnutrition rather than to an individual disaster, and specifically nutritional aspects are emphasised in the interventions, but the basic
approach is risk management, seeking to limit the likelihood of future
disaster. Measures to combat chronic food insecurity are an important
element of anti-famine initiatives.
257
slow-onset
disasters
8/3/04
15
Good Practice 4th

46. Good Practice 4th
8/3/04
2:44 pm
Page 258
disaster risk reduction: mitigation and preparedness
Table 15.1 Interventions against malnutrition in Tanzania
Problem
Possible causes
Programme interventions
1. Inadequate food in
households (most severe
a few months before the
harvest)
• lack of household
planning
• choice of wrong crops
• failure of rains
• poor crop management
• storage losses
• shortage of agricultural
inputs
• lack of income to buy
food
• training of trainers in
household food planning
• promotion of droughtresistant crops
• improved storage
2. Inadequate nutrient
intake (especially in
children)
•
•
•
•
poor economic resources
nutritionally poor diet
shortage of fuelwood
shortage of fruits and
vegetables
• scarcity of meat
• scarcity of beans and
other legumes
• too much workload for
mothers
• promotion of
income-generating
activities
• nutrition education,
especially to mothers,
through health workers
• training and other inputs
for village forestation
and home gardening
• training and inputs for
small animal keeping
• promotion of grain
milling and appropriate
technology
3. Lack of awareness of
good weaning practices
• lack of awareness of
children’s nutrient needs
• inadequate feeding
frequency
• scarcity of energy-dense
foods (e.g. groundnuts,
cooking oil)
• dietary bulk
• provide weaning recipes
based on local foods
• mobilise communities to
provide extra food at
child care posts
• campaign on use of
kimea (flour with high
nutritional value)
M. Hubbard, Improving Food Security: A Guide for Rural Development Managers (London: IT
Publications, 1995), pp. 30–31.
Interventions should be linked as far as possible to the coping strategies that
households use. As affected people are likely to employ a variety of strategies, and may alter the type and mix of coping methods in use quite rapidly,
agencies need to be flexible in their approach.
258

47. 2:44 pm
Page 259
chapter 15 slow-onset disasters
Some of the components that might make up an integrated approach are
outlined in the following sections.
15.3.5 Protecting food production
Drought mitigation and food security projects often used to focus on irrigation or soil and water conservation techniques. Nowadays it is usual to
attempt a broader mix of activities that support all aspects of agricultural
production, but irrigation and soil and water conservation remain important
elements of such work.
There are many different methods of protecting the soil against erosion from
wind or water and preventing water run-off. These include terracing, digging
furrows and ridges, planting grasses, bushes and trees, building walls of
stone or earth, planting in holes and pits, and mulching. The most appropriate method for each location depends upon its physical features (the
nature of the soil, the terrain and climatic conditions) and local capacities
(the materials, skills and other resources available).
Intercropping (mixing different crops in the same plot) is a traditional coping
strategy that has been undermined in some places by the advance of monoculture. Where the crops chosen differ in their resilience to drought, diseases
or pests, intercropping is a way of spreading risk, since it increases the probability that some crops will survive. It can also be beneficial to growth: some
plants give shelter or shade to others, or provide nutrients to the soil. Other
agricultural practices that can be encouraged include alternative systems of
crop rotation, manuring and composting. Integrated pest management,
based on intercropping and the use of insect-repelling plants and crops or
pesticides (including those made from traditional recipes using local plants),
is another feature of many successful food security initiatives.
Farming should be considered as a process. In real life, agricultural methods
are not fixed; the type and mix of crops vary. Farmers do experiment, if the risks
of doing so are not too great. Supporting organisations should therefore aim to
give food-insecure communities more livelihood options and greater flexibility.
Initiatives that begin as narrow technical interventions often find that they have
to widen their scope in order to become more effective, as well as sustainable.
Government agricultural extension services and some NGOs have promoted
new, hybrid versions of staple food crops at the expense of traditional varieties. The newer varieties give higher yields, but usually only in favourable
conditions with ample water and fertilisers. Traditional varieties tend to be
more resilient to environmental stresses such as drought. This alone should
259
slow-onset
disasters
8/3/04
15
Good Practice 4th

48. Good Practice 4th
8/3/04
2:44 pm
Page 260
disaster risk reduction: mitigation and preparedness
be a sufficient caution against using newer crops in drought-prone areas, but
their heavy – in some cases, coercive – promotion has also threatened biodiversity and undermined traditional knowledge of alternative varieties.
Research in the past 10–15 years has uncovered extensive indigenous knowledge of agricultural plants and how to grow them.10 Many food security
projects have found that, by protecting and sharing such knowledge, and the
traditional seed varieties concerned, they widen the options open to communities and increase their resilience to hazards such as drought. Seed banks,
fairs and demonstration plots are effective methods of preserving, promoting
and sharing (e.g. Case Study 15.2).
Case Study 15.2
Improving food security in a drought-prone area
Chivi District, in Zimbabwe’s Masvingo
Province, is an area of poor soils and
low rainfall. In 1990, the Intermediate
Technology Development Group
(ITDG) began to explore methods of
working with communities and local
organisations in the district’s Ward 21
to improve food security among the
1,300 households there. This
coincided with a period of drought
that killed many cattle in Chivi.
ITDG intended a strongly
participatory approach from the
start, but still had to overcome
considerable local cynicism about
the value of outside agencies, while
many villagers were puzzled that it
had not come with a pre-formulated
programme and was not offering
financial or material support.
An extensive assessment process
carried out with the community
identified a wide range of problems in
producing food and sustaining
livelihoods. Only after this period of
consultation, which took six months,
did the project begin to set priorities
and make plans – again involving the
community, many of whom had never
gone through a participatory process
of this kind. The project worked with
two main local institutions, farmers’
clubs and women’s garden groups, as
these were most directly involved in
food production.
The first phase of implementation,
from 1992–95, addressed three needs:
water for fields and gardens, pest and
disease control, and fencing to protect
gardens from animals. At the same
time, over 1,800 community members
took part in training to improve their
capacity to identify problems and
solutions, communicate, and manage
their own organisations.
(continued)
260

49. 8/3/04
2:44 pm
Page 261
15
chapter 15 slow-onset disasters
slow-onset
disasters
Good Practice 4th
Case Study 15.2 (continued)
During the first two years, a range of
technologies were tried out by the
farmers and gardeners through
experimentation in their own fields.
Those found to be most effective
were quickly taken up by other
community members. They included
sub-surface irrigation using clay
pipes, pots and bottles; terracing,
rock catchments, ‘tied’ ridges,
infiltration pits, mulching and the use
of underground plastic sheeting to
increase water retention; digging and
improving wells; winter ploughing,
intercropping, and the use of termite
soil as a fertilizer and moisture
retainer; growing ‘live’ fences of sisal
and introducing two machines to
make wire fencing at half the price of
ready-made products.
Indigenous knowledge, often
disregarded by outside experts, was
used wherever possible. Seed fairs
were held to revive local crop
varieties, share information on them
and demonstrate their value.
Surveys and discussions showed
that many farmers knew of local
plants that could be used to make
effective pesticides, but were
reluctant to tell others because they
feared their knowledge would be
thought old-fashioned in an age of
modern chemical pesticides. The
project made sure that their
knowledge was recorded and
shared.
261
All of these methods were widely
adopted and effective in boosting
production. However, as the project
progressed it was clear that it needed
to pay more attention to marketing.
Women gardeners carried out surveys
of demand and prices for various
vegetables, and as a result the groups
began diversifying their production to
meet these opportunities.
As the project took off, with
widespread adoption of the various
techniques and growing numbers of
local people taking part, it also
expanded into other wards. Reviews
and evaluations over the years have
found increased and more reliable
crop yields and a steady growth in
the variety of crops being grown.
Buyers began bringing trucks into
the area to buy up surpluses, and
women began sending their
husbands to neighbouring areas to
sell the produce from their vegetable
gardens. Some garden groups set up
a revolving loan fund, and
community organisations were seen
to be far more confident not only in
managing their own affairs, but also
in their relationships with outside
agencies and government
agricultural extension workers.
K. Murwira et al., Beating Hunger: The
Chivi Experience. A Community-based
Approach to Food Security in Zimbabwe
(London: IT Publications, 2000).

50. Good Practice 4th
8/3/04
2:44 pm
Page 262
disaster risk reduction: mitigation and preparedness
Food production is inseparable from issues of access to land and land ownership. Many of the problems faced by poor communities in drought-prone
areas arise from unequal distribution of land and natural resources, which is
the product of historical, political and socio-economic forces. Some developing-country governments have attempted land redistribution, but this
approach has fallen by the wayside with the current dominance of market-led
economic policy. On the other hand, the protection of common property
resources such as forests and grazing lands is definitely a ‘live’ issue, for
which determined advocacy may be needed.
15.3.6 Preserving food: crop and seed stores and banks
Many drought mitigation projects include crop and seed storage among their
interventions, and there is now plenty of guidance on this subject.11
Storage falls into two main categories: household stores and community
grain and seed banks. Where people are poor and agricultural output is low
or highly vulnerable to climatic variations, it is important to maximise crop
preservation. Inadequate storage can lead to crops rotting, becoming
diseased or contaminated, or being eaten by pests. The amount of grain and
seed lost because of this varies greatly according to location, but in many
cases may be as much as a third of the crop.
Crop stores and seed banks help to ensure that there is food to eat during the
lean season, and that there are seeds to plant. They offer security against
rising food prices during the hungry period. By storing a wide variety of local
seeds, they maintain biodiversity. They can also protect crops and seeds
against other natural hazards such as floods.
Appropriate technologies come to the fore here. Low-cost techniques and
materials can often be used to make stores, while in some instances traditional knowledge and methods can be adopted, or adapted. Examples of
appropriate technologies include sealed clay pots, baskets lined with clay or
plaster, plastic sheeting, sacks, metal bins and some forms of underground
storage such as lined and covered pits. Stores can be raised above the
ground on wooden poles to protect them against rats and mice (with guards
made of old tin cans on the poles to prevent the animals from climbing up).
Certain varieties of crop and seed may be more resistant to pests and disease
than others, and it is useful to explore traditional knowledge of this.
Traditional pesticides such as ash, some types of edible oil, and certain local
plants may preserve crops against insect attack. Some crops can be
262

51. 2:44 pm
Page 263
chapter 15 slow-onset disasters
preserved for longer by drying or smoking them. In other cases crop
processing – into flour, oil, jams or pickles, for instance – is an effective
preservative, as well as creating a product for sale. Exchange visits enable
farmers to see different storage and preserving practices and discuss their
effectiveness.
A few words of caution are needed, nevertheless. First, it is important to
identify whether crop losses in storage result from poor storage itself, or from
harvesting and post-harvest preservation practices. If the latter are inadequate and introduce disease or contamination into the crops, good storage
systems will make little difference.
Second, community seed or grain banks present storage and management
challenges that are quite different from domestic stores. The storage challenges are technical, arising from the scale on which produce has to be
stored, but in many cases similar technologies to those employed by households can be used.The real challenge is management. A seed or grain bank is
a bank, not merely a store. These banks can be run in different ways but the
principles are standard: usually they buy grain from their members and sell it
back at below market rates, or they run as savings and credit schemes.
Procedures governing how households deposit seeds or grain with the
stores, how to sell or lend seeds and grain back to them, and how to deal
with defaulters must be worked out carefully, and they must be transparent.
Grain/seed banks must be planned with communities, built by them and
managed by them – crucially, they must be run on behalf of the whole community. This may require training in organisational development, literacy and
accounting procedures. Projects dominated by outside agencies are prone to
failure. However, start-up funding may be needed to build stores and purchase
initial grain stocks, and top-up funds or grain may be necessary if a drought hits
before the scheme is securely established or at times of severe crisis. Any
outside organisation attempting to introduce such schemes needs a high level
of skill and experience in community development, in the broadest sense, and
if it lacks this should bring in the relevant expertise. Moreover, there has to be a
high level of trust between the community and the development or disaster
organisation that is helping it, which may take years to build.
A further important point is that community seed or grain banks should be
part of an integrated food security or rural development programme. On their
own, they cannot provide complete food security. Other factors affecting
success are outlined in Case Study 15.3.
263
slow-onset
disasters
8/3/04
15
Good Practice 4th

52. Good Practice 4th
8/3/04
2:44 pm
Page 264
disaster risk reduction: mitigation and preparedness
Case Study 15.3
Cereal banks in Burkina Faso
Cereal banks were introduced to
Burkina Faso in the mid-1970s. By
1986, there were an estimated 1,177
banks supported by government and
NGO programmes. They had three
basic elements in common: a storage
facility; a ‘rotating fund’ in grain or
cash that allowed the bank to buy,
sell or lend grain; and a managing
committee chosen by villagers.
drought in the mid-1980s greatly
reduced farmers’ ability to repay the
bank for borrowed grain); the amount
of time and effort needed to manage
schemes; finding adequate ways of
compensating the banks’ managers
for their time; and tension when
people were refused credit. The need
to constantly remind villagers about
how credit works was noted.
In a 1987 study of nine banks
supported by three NGOs, villagers
were asked to identify their
achievements and weaknesses.
Among the benefits were: a measure
of food security even during
prolonged drought; reducing
drought-induced emigration;
savings in time because villagers no
longer had to go to market to buy
grain; enabling farmers to escape
the high prices imposed by grain
merchants; improved management
skills that could be applied to other
local projects and stimulating
mobilisation of community members
for such initiatives; facilitating
emergency relief assistance; and
generating funds for other
community activities. The bank
meetings often served as a forum to
discuss other village matters.
The study concluded that the cereal
banks were best suited to dealing
with yearly grain shortfalls during
the planting and growing periods,
and highlighted the increased
organisational capacity of villages
resulting from the initiatives. There
were more questions about the
banks’ ability to respond to
prolonged drought and famine. They
were able to blunt the impact of
drought and mitigate or even prevent
famine. However, they risked their
long-term financial stability by doing
so since farmers were often unable
to reimburse them adequately after
the crisis had passed. To protect
them against this risk, special
donations of grain or funds were
needed to provide emergency
supplies or replenish stocks.
Difficulties included: maintaining high
rates of reimbursement (in several
locations the pressure of a severe
264
M. B. Anderson and P. J. Woodrow, Rising
from the Ashes: Development Strategies
in Times of Disaster (London: IT
Publications, 1998), pp. 185–205.

53. 8/3/04
2:44 pm
Page 265
15
chapter 15 slow-onset disasters
slow-onset
disasters
Good Practice 4th
15.3.7 Preserving water supplies
Water shortages affect crops, livestock and people. Town-dwellers rely on
large-scale water infrastructure. In the countryside, communities may have
access to a variety of sources: rivers, ponds, wells and small dams. The
extent of access depends on distance, ownership of the water resource, and
the cost and technical difficulty of collecting or extracting water.
During a drought, rural communities rely on local water sources, or move out
of the area altogether. In some emergencies, governments and aid agencies
may use tankers, but these are a costly, short-term response to the problem
and are unlikely to reach the most remote communities.
There are two main options for improving water supplies. The first is to
improve access to underground water sources, for instance by deepening
wells or digging new boreholes. Such measures may be beyond the financial
resources of many poor communities. If the water is fed into irrigation
schemes, then the cost of installing and maintaining irrigation pipes and
channels must be added, although in many locations these may be able to
connect to traditional irrigation networks.
The second method is ‘rainwater harvesting’. There are many different ways
of harvesting rainwater for agricultural and domestic use. They include:
• building water-storing dams and percolation dams (dams that slow the
rate of rainwater run-off and so increase absorption into the soil, thereby
recharging local groundwater);
• building community or domestic storage tanks;
• lining ponds with plastic to improve water retention; and
• putting up stone or earth bunds to improve absorption and reduce soil
loss (brushwood and strips of grass or other plants can also be used).
Many of these methods are traditional, and hence the knowledge and skills
needed to build and maintain them are present in the community. In other
cases, the relevant technical expertise is easily acquired (see Case Study
10.6, page 160).12
Most rainwater harvesting methods are cheap compared to digging wells and
pumping water from more remote sources. Communities can provide labour
and in some cases building materials – bunds and some dams use just stones
and earth. Many readily available materials can be used to catch, channel and
store rainwater. A recent inventory of materials employed in Sri Lanka for
265

54. Good Practice 4th
8/3/04
2:44 pm
Page 266
disaster risk reduction: mitigation and preparedness
domestic rainwater harvesting listed tin sheets, palm leaves, plastic sheets,
the stems of plants such as bananas and bamboos, tree trunks and rock
cavities, as well as more conventional gutter pipes and tanks.13 Nevertheless,
the cost to poor people is not always small (see Case Study 10.3, page 152).
Even where freely available materials are used and labour is voluntary, the
task of constructing larger-scale structures such as tanks and dams is
substantial. Collective action is needed in such cases.
Rainwater harvesting can be highly effective. India, where there has been a
considerable revival of traditional methods in the past decade, appears to have
had considerable success. The expansion of rainwater harvesting there was
assisted by the existence of centuries-old traditions and techniques, coupled
with technical support and vigorous advocacy from Indian NGOs. It was further
stimulated by prolonged drought. Case Study 15.4 illustrates the potential of
rainwater harvesting, and the challenges to implementing successful schemes.
The choice of approach to water provision varies according to location, and
may vary over even a small area according to such factors as the topography,
the level of dependence on irrigation compared to rainfall, and the moistureretaining capacity of different soils.
In some cases, the problem may be one of access to water, not its physical
availability. Equitable water distribution is the goal. Communities in droughtprone areas often have sophisticated systems for this. Local management
structures should be reinforced where necessary.
15.3.8 Preserving livestock
In many remote areas, poor people’s livelihoods depend on livestock. This is
most obviously true in the case of nomadic pastoralists. However, farmers
may also rely heavily on livestock if they work arid or hilly land that does not
support intensive crop farming. Livestock are a valuable asset, providing
food, income and agricultural inputs (manure, pulling ploughs and carts).
Herds grow as new animals are born. They can be moved easily and looked
after by children.
Some livestock, such as goats and camels, are good at withstanding water
shortages. A one-year drought may have little effect on the size of a herd
since the animals can be moved. But when the drought is prolonged, poor
people are often forced to sell animals to raise money for food (usually,
breeding animals are kept and others sold). If this happens on any scale, it
drives livestock prices down, generally at the same time as grain prices are
266

55. 8/3/04
2:44 pm
Page 267
15
chapter 15 slow-onset disasters
Case Study 15.4
Rainwater harvesting in Kenya
A severe famine in 1979–80 in
Turkana District in north-west Kenya
led to a number of long-term
development initiatives. Several
agencies became interested in
rainwater harvesting as a way of
supporting the Turkana people, who,
though essentially pastoralists, had
traditionally grown crops, principally
sorghum, during the wet season.
Many of the early schemes did not
work well, for both technical and
operational reasons, and in 1985 a
group of NGOs began collaborating
on a project in Lokitaung Division to
test alternative approaches, based
on simple construction and
surveying technologies.
Local people were soon able to
survey sites, and design and build
water-retaining bunds of earth
without assistance, but it took time
to overcome other technical
challenges. For example, although
many rainwater harvesting methods
were known, there was no
consensus about the most
appropriate techniques in particular
geographical and socio-economic
contexts. It was difficult to build
structures that would hold enough
water to flood a garden site in a year
of poor rains, but that would also not
be swept away by the pressure of
floodwater when rains were heavy.
Digging water-retaining earthworks
is time-consuming and labourintensive. The project promoted the
use of animals (especially donkeys
and oxen) to help move earth, but
had to develop suitable scoops and
harnesses, and overcome resistance
from herders who refused to let their
own animals be used. Later, less
labour-intensive methods such as
stone lines and grass strips were
introduced.
At the same time, understanding of
the role of the sorghum gardens
deepened. It was realised that,
although rainwater harvesting could
improve sorghum yields in some
years, it could not provide the whole
livelihood for any group of local
people, even if they wished to give
up their nomadic way of life.
Improved gardens were now viewed
not as a means of support for people
made destitute by the earlier famine,
but as a means of complementing
existing pastoral livelihoods.
The project was handed over to local
management in 1988, and a sister
project in Kakuma Division was set
up in 1989. The two projects support
garden improvement through
provision of local technicians
(continued)
267
slow-onset
disasters
Good Practice 4th

56. Good Practice 4th
8/3/04
2:44 pm
Page 268
disaster risk reduction: mitigation and preparedness
Case Study 15.4 (continued)
(fundis) to design structures, the
sale or loan of tools, seed and grain
storage facilities, and technical
assistance for repairs.
In 1997, an evaluation of the
Lokitaung and Kakuma initiatives
observed that more than 2,000
people were involved in rainwater
harvesting in the two projects, with
340 gardens covering 170 acres.
There was a gradual increase in the
number of gardens in most locations,
even though food-for-work had long
since ceased to be available for
labour on construction. In years of
medium and poor rainfall, the
improved gardens increased sorghum
yields by up to four times more than
traditional plots. The reliability of
crops was also much greater.
Most of the work on the gardens was
carried out by women, but they had
relatively tight control over the use
of the harvest. The sorghum grown
was used mostly to boost household
food supplies, to buy goats and as
an investment in social networks (i.e.
through giving or lending to other
members of the community).
A. Cullis and A. Pacey, A Development
Dialogue: Rainwater Harvesting in
Turkana (London: IT Publications, 1992);
C. Watson and B. Ndung’u, ‘Rainwater
Harvesting in Turkana: An Evaluation of
Impact and Sustainability’, unpublished
report (Nairobi: Intermediate
Technology Development Group Kenya,
1997).
rising with scarcity. This can result in the severe depletion of herds, from
which it may take many years to recover.
Agency interventions to protect livestock-dependent communities during
drought crises are generally of the following main kinds:14
• Increasing grain and fodder supplies to the area (the former to help keep
grain prices down, the latter to keep animals fed).
• Removing surplus animals (e.g. by buying, slaughtering and processing
them).
• Giving broader support to communities’ livelihoods, so that they do not
have to sell their breeding animals.
• Where herds are severely depleted and their owners destitute, restocking
with new animals.
268

57. 8/3/04
2:44 pm
Page 269
15
chapter 15 slow-onset disasters
Case Study 15.5
Pastoralists’ coping strategies
The Gabra are a pastoralist tribe
herding camels and goats over arid
lands in northern Kenya and
southern Ethiopia. Their economy
depends on animal products. Like
many other pastoral groups in the
region, the Gabra were badly hit by
drought and animal disease in 1991.
Many animals died.
Owing to the severity of the crisis,
community meetings were held to
agree collective coping strategies.
Rich families agreed to give goats to
the poorest and to send them milk
daily; meat from slaughtered camels
was also distributed periodically.
Traditional reciprocal relationships
with another pastoral group, the
Borana, were invoked because of the
emergency: the Borana were asked
to give cattle.
In some cases, rich relatives and
friends from other districts loaned
animals to replace those lost to
disease, until the herd could be built
up again. In other cases, herders
travelled far to meet relatives and
friends and offer services in
exchange for animals. Some children
were sent to live with distant
relations who were less badly hit.
B. J. Linquist, ‘Caring for the Poor, Gabrastyle: Indigenous Relief and Development
Strategies’, Appropriate Technology, vol.
21, no. 4, 1995, pp. 6–8.
Most of these interventions are closer to the emergency response end of the
disaster management spectrum than to disaster preparedness. Long-term
livelihood support is needed to give greater security.
Pastoralists in particular have long-established methods of coping with
drought, based largely on moving animals to other areas and partly on livestock sales (see Case Study 15.5). But with traditional grazing lands increasingly under threat from privatisation for ranching or other forms of
development, coupled with conflict in African countries in particular, it is
becoming harder for them to put these coping strategies into practice.
Where communities are heavily dependent on livestock that cannot easily be
herded elsewhere (e.g. dairy cows), collective schemes might be established
to maintain fodder supplies during droughts through bulk purchase and
269
slow-onset
disasters
Good Practice 4th

58. Good Practice 4th
8/3/04
2:44 pm
Page 270
disaster risk reduction: mitigation and preparedness
Case Study 15.6
Supplying food to remote villages
In the dry rural district of
Banaskantha in Gujarat, SEWA and
the Disaster Mitigation Institute
(DMI) set up a programme in 1993 to
purchase basic foods wholesale in
local markets. The food was given to
the poorest members of SEWA’s local
groups, who paid for it then or later
out of the money they earned from
their involvement in other incomegenerating activities supported by
SEWA.
The scheme, known as Shakti Packet,
was targeted at remote villages where
government distribution systems were
absent or ineffective. One of its aims
was to reduce women’s indebtedness
to shopkeepers. Another was to give
them greater choice of foods. SEWA’s
local women’s groups chose the
products to be purchased. These
included millet, rice, pulses, sugar,
edible oil, spices, tea, salt, sugar and
green vegetables, as well as non-food
items such as soap and even
contraceptives.
Early results from five villages where
the Shakti Packet scheme was
piloted showed that it was reducing
women’s indebtedness and allowing
children to eat green vegetables
more regularly. The initiative was
seen as complementary to SEWA’s
wider development programme for
poor women in the district, which
included savings groups, dairy
cooperatives and other incomegenerating activities. Challenges to
expanding the scheme were the
rising price of food in the markets,
limited working capital at village
level, and difficulties in linking it with
the government’s public distribution
system.
M. R. Bhatt, ‘Shakti Packet: Increasing
Food Security and Nutrition of Rural
Women’, paper presented to the conference ‘Gender Approaches to Health in
Emergency Situations’, San Marino, 8–9
December 1995.
community-managed plantations. This has worked effectively in parts of
India, but is likely to need strong institutional support.15
15.3.9 Food aid
Food aid is generally regarded as a matter for governments and international
agencies, because only they have the capacity to purchase and distribute
270

59. 2:44 pm
Page 271
chapter 15 slow-onset disasters
large volumes. It is possible to set up more local schemes, as Case Study 15.6
shows, but the challenges are considerable.
15.3.10 Nutrition
Being able to monitor and combat malnutrition is an important component of
all food security work, including early warning. Assessment of nutritional
deficiencies requires specialist expertise and is beyond the scope of this
book, but several major agencies have produced technical guidelines for the
management of nutrition in crises.16
Advances in nutritional assessment methods have increased agencies’ ability
to monitor and manage crises as they enter the acute stage. Since the early
1990s, nutritionists have widened the focus of their work from malnourished
individuals to larger populations, and from a narrow set of technical interventions to combat malnutrition to a broader range of strategies, policies and
programmes that take account of related causal factors, such as water, sanitation, health and social care.
Multi-sectoral approaches are ideal in theory, but their complexity causes
problems. In practice, nutrition initiatives often have to make the difficult
choice between concentrating their resources on the direct alleviation of
malnutrition – usually through feeding programmes – or addressing some of
its underlying causes.
15.3.11 Protecting livelihoods
Where support to livelihoods is concerned, many of the steps that organisations will need to take in drought-prone communities can be regarded as
general development interventions just as much as disaster mitigation.
A broad range of interventions will be needed to stimulate local economies so
that poor people are no longer so dependent on agriculture for their food and
income. Economic diversification is the key to success, through on- and offfarm enterprises. Where a family’s income relies on a range of different
economic activities, there is a greater degree of protection against the failure
of any one of them (e.g. harvest failure due to drought).
Diversifying agricultural production is part of this. Many food security projects
encourage farmers to grow a wider range of crops, establish kitchen gardens
and orchards, keep poultry or set up fish ponds. This has the twin objectives of
improving food supplies and generating produce that can be sold. Technical
271
slow-onset
disasters
8/3/04
15
Good Practice 4th

60. Good Practice 4th
8/3/04
2:44 pm
Page 272
disaster risk reduction: mitigation and preparedness
assistance, if needed, must be supplemented by training in business skills,
marketing and organising production to meet market demands.
Many drought mitigation projects, in Africa particularly, support household
gardens in addition to farms. Gardens are typically used to grow vegetables
that will give a more varied diet and can be sold. In many communities they
are managed by women, who thereby gain more control over household food
supplies and income.
Another common approach to enhancing livelihoods is by processing agricultural products – for example grain milling, oil processing, making jam and
peanut butter, or fruit and vegetable drying. Processing preserves crops, and
often adds value.
Off-farm employment may be regarded as more secure against drought than
agriculture, but this impression can be misleading. Local industries that
depend on water or agricultural products are also vulnerable: for instance, a
grain mill stops working when there is no more grain to mill. In such cases,
the impact of the drought may be delayed until some time after crops have
failed, but this is only a delay. It is therefore common for people to migrate
well away from their communities in search of work during a drought. In
western India, for example, village men head for brickworks and building
sites in the cities during prolonged droughts. The collapse of the agricultural
economy also affects rural people’s purchasing power, and so has a knock-on
effect on shopkeepers and traders supplying goods.
Plans for local economic diversification must take hazards into account and,
where possible, find productive activities that are unlikely to be directly
affected. Artisanal crafts may be suitable, as the supply of raw materials may
not be hit by a drought or other hazard and the markets may be some distance
from the affected area. Creating alternative enterprises is a complex task, and
likely to require support in technical and business skills, credit and market
access. Specialist assistance should be sought to carry out such work.
The degree of complexity involved is illustrated by one aspect of economic
diversification: the need for well-functioning local markets. These make it
easier for vulnerable people to buy and sell at times of need and, by distributing efficiently and moderating shortages, help to keep down the prices of
essential items such as food. Development and disaster planners will
probably need to act in several ways to strengthen local markets and improve
poor producers’ access to them, by improving local transport infrastructure,
helping to disseminate information about prices, lobbying against damaging
272

61. 2:44 pm
Page 273
chapter 15 slow-onset disasters
Case Study 15.7
From food security to community empowerment
The Kebkabiya district in Darfur,
western Sudan, was badly affected by
the drought and famine of 1984–85.
In 1985, Oxfam began a post-disaster
initiative to improve food security. In
the first phase of the project, 12
community seed banks were
established to serve a community of
30,000. The second phase, which
began in 1989, introduced other
measures to improve food
production: animal health care, the
use of animals for ploughing, pest
control, soil and water conservation
and community development.
The project embodied community
participation from the beginning, but
its nature changed over the years. In
the first phase, the seed banks were
managed by local committees, almost
exclusively male, nominated by
existing village authorities. There was,
however, extensive consultation with
communities in planning and
establishing the seed bank scheme. As
the seed bank committees gained in
confidence, they took more decisions
on their own, and when the second
phase of work was prepared in 1989
there was no need for Oxfam to be
involved in day-to-day management.
The committees also began taking on
extra tasks, such as organising
training for farmers and supervising
revolving funds for buying seeds.
273
For the second phase, new
structures were needed to manage
this wider range of activities. These
were far more democratic, with each
village in a group of 5–12 electing
one man and one woman to
represent it on a Village Centre
Committee. In turn, each of the 16
Village Centre Committees elected
one man and one woman to
represent it on the Project
Management Committee which
oversaw the whole project. In 1990,
the Project Management Committee
registered as a separate NGO, the
Kebkabiya Smallholders’ Charitable
Society, to take over the project from
Oxfam. This is a membership
organisation, with each household
paying a small fee. Members elect
representative committees at
different levels.
The project was always staffed
almost entirely by Sudanese, but a
handover to the communities had
been envisaged from the beginning.
The approach was gradual,
transferring authority over several
years, and Oxfam continued to
provide technical back-up when
required.
P. Strachan and C. Peters, Empowering
Communities: A Casebook from West
Sudan (Oxford: Oxfam, 1997).
slow-onset
disasters
8/3/04
15
Good Practice 4th

62. Good Practice 4th
8/3/04
2:44 pm
Page 274
disaster risk reduction: mitigation and preparedness
market restrictions, promoting more efficient methods of storing, preserving
and transporting perishable goods such as food products, supporting smallscale decentralised processing facilities (e.g. grain milling), and providing
credit and training in small enterprise management.17
While communities with high incomes generally suffer less from malnutrition,
the link between wealth and nutrition is not clear-cut: there can be considerable differences within communities and households. Other factors play an
important role. For example, the level of education among women has a great
influence on dietary, hygiene and health practices. The benefits of increased
income also depend on who earns and controls the money.
Maintaining natural resources such as forests, grazing land and sources of
water is important to food security, especially where these are held in
common. More intensive use of common property – for grazing, collecting
wild food and roots, or fishing – is an important coping strategy when there is
a drought. Preserving these resources against encroachment by private interests or their destruction by alternative forms of commercial or state-sponsored development, such as farming, logging and the construction of dams, is
usually difficult. Local voices are unlikely to be heard unless their campaigns
are supported by organisations with resources and lobbying skills.
Reforestation to mitigate drought or other hazards presents significant challenges. Communities need to be convinced of the need for it, and must have
strong incentives for investing in trees that may not produce anything of
economic value for many years. Expertise in forest management is essential, and
should be brought in from outside where necessary. There are many anecdotal
accounts of tree-planting initiatives as part of disaster mitigation programmes
(protecting land erosion or providing wood for disaster-resistant housing) that
failed because project managers lacked experience of this kind of work.
15.4 Monitoring and warning systems
There are now many kinds of system warning of food shortage. The first
major one was the FAO’s Global Information and Early Warning System
(GIEWS), developed after the Sahel famines of the early 1970s to monitor
food production and supplies at national level and assess emergency food
needs in areas facing critical shortages. The African famines of the mid-1980s
showed the need for improved warning systems, and several more were
established in the Sahel and the Horn of Africa between 1985 and 1990.
These usually combined hazard/meteorological monitoring (see Box 15.1)
and assessments of food production levels after the harvest season.
274

63. 8/3/04
2:44 pm
Page 275
15
chapter 15 slow-onset disasters
slow-onset
disasters
Good Practice 4th
Box 15.1
Drought monitoring
Drought monitoring systems look at
two main indicators: rainfall and
vegetation. The former is monitored
by extensive networks of rain
gauges, the latter mostly through
remote sensing by satellites. In both
cases, large amounts of high-quality
data can be generated. These can be
supplemented by other
meteorological information (such as
rainfall forecasts) and hydrological
data (such as monitoring of
groundwater supplies and the level
of water tables).
Although rain gauges are relatively
simple technologically, and collecting
data often depends on local staff or
volunteers, rainfall monitoring needs
to take place on a large scale if it is
to be of value in assessing overall
needs and particular priorities.
Management of such systems is
generally taken on by government
meteorological services, and feeds
into the well-established and
effective national and international
meteorological information systems.
Moreover, there is growing use of
rain gauges that transmit data
automatically to distant monitoring
stations, using radio signals or other
electronic means of communication
(this is called telemetry). In the case
of remote sensing of vegetation,
which is the only really effective way
of compiling information covering a
wide area, the cost of procuring and
analysing satellite data is so high
that this task too is generally left to
international and government
services.
Monitoring of rainfall and vegetation
does not indicate how much food
people have, or need. Remote
sensing does not distinguish
between different kinds of
vegetation, so does not show how
well crops are growing. Since
different food crops vary in their
levels of tolerance to drought,
rainfall monitoring is of limited value
as an indicator of the availability of
food. It is for this reason that food
security or famine information and
warning systems have become an
increasingly important tool for
disaster managers during the past
20 years.
Because they operate on national and regional scales, such systems are best
managed by governments and international organisations. They are also
geared to large-scale disasters requiring international aid.
275

64. Good Practice 4th
8/3/04
2:44 pm
Page 276
disaster risk reduction: mitigation and preparedness
However, local early-warning systems have an important role to play, and
many have been established since the mid-1980s, often by NGOs. Many
researchers and practitioners consider them to be particularly successful in
monitoring impending food crises. They tend to draw on a wider range of indicators of food and livelihood insecurity than the larger systems. They also
rely far more on qualitative data, and involve higher levels of community
participation. They are better able to take account of local variations in food
security, and are more sensitive to local coping strategies and vulnerability.
They can recommend appropriate interventions to local decision-makers,
who will have a better understanding of conditions on the ground and a
greater sense of urgency in responding to problems. They are easier to
manage than large-scale, centralised systems, but they tend to suffer from a
lack of skilled personnel (see below) and are open to manipulation by
powerful local interests.
Systems of all kinds have shifted their emphasis from the simple availability
of food to considering which groups do not have access to food: this takes
them logically into vulnerability and livelihoods analysis. As a result of these
developments, the targeting of food aid has improved considerably.
The effectiveness of early-warning systems varies in practice. There are four
main reasons for this:
1. The nature of the system itself and the information provided – the range
of indicators used, accuracy of the data, the timeliness of warnings.
2. The institutional context within which the system is located, and institutional links to decision-makers.
3. The broader political environment. Decisions about when and how to
intervene are political, and therefore influenced by many other factors.
4. Logistical obstacles to launching a timely and adequate response.
Most analysis of systems concentrates on the first and fourth of these
factors, but research into responses to the African droughts of 1990–93 has
shown that reasons 2 and 3 are most important in explaining if early-warning
information is used, and variations in performance between different warning
systems. Early-warning systems did sound the alarm about impending food
crises, but the response systems failed to act early enough.18 Conventional
systems are also much less effective in conflict-induced famine.
Early-warning systems take many forms in their institutional set-up and
location, the resources available to run them, and the information that they
collect and process. However, all are designed to stimulate action by
276

65. 2:44 pm
Page 277
chapter 15 slow-onset disasters
informing decision-makers about food security conditions and people’s
needs. As information systems, their functioning is theoretically straightforward (see Figure 15.1), but in practice there are many obstacles.
An efficient, effective early-warning system for drought-related famine should
have the following three characteristics:19
1. It should be capable not only of warning of large-scale famine, but also
should be sensitive to changes in food-security status before famine
threatens, and able to detect localised pockets of acute food stress.
2. It should generate a response that provides assistance early in the
‘famine spiral’, before families and communities are reduced to destitution.
3. It should stimulate interventions that protect livelihoods before lives are
threatened. This implies providing a wider range of relief than food aid, as
well as a more developmental approach.
The rest of this section focuses on two aspects of warning systems for food
insecurity and famine: the selection and collection of indicators, and the
management of local-level systems. The discussion is necessarily brief, as
this is a complex technical subject on which a great deal has been written
elsewhere.20
15.4.1 Data and indicators
Under the influence of Sen’s ‘entitlement’ theory, food security information
and famine early-warning systems have increasingly incorporated a wider
range of indicators of the availability of food and the ability to procure it. This
includes data on the market price of food and other essential goods, family
and community behaviour (the adoption of particular coping strategies) and
the availability of employment opportunities in the area, as well as more
conventional data on rainfall and levels of groundwater, crop production
(surveys before and after the harvest), nutritional status and food supplies.
However, many early-warning systems still have too narrow a focus. They rely
largely on indicators of food production and supply, instead of indicators of
access to food. These are easier to collect and are often believed to be more
accurate, though this may not be true in many cases. Socio-economic indicators, which tend to be harder to collect and draw more on qualitative data,
are less influential in overall decision-making. They are more likely to be used
in identifying those most in need of food aid. Coping strategies, which are
difficult to monitor and interpret accurately, are rarely incorporated systemat-
277
slow-onset
disasters
8/3/04
15
Good Practice 4th

66. Good Practice 4th
8/3/04
2:44 pm
Page 278
disaster risk reduction: mitigation and preparedness
Figure 15.1
An early-warning system as information system
Financial and
material inputs
Environment
Diverse sources of
information
Feedback
Data
EWS
(as organisation)
Boundary
Observation and
measurement
Observational
filter
Information
Problem
definition
Synthesis
of data
Analysis of options
(recommendations)
Users of
information
Knowledge
Decision-making
Donors
Government
Other
sources of
information
Government
Implementation
NGOs
(donors)
Feedback
Beneficiaries
M. Buchanan-Smith and S. Davies, Famine Early Warning and Response – The Missing Link
(London: IT Publications, 1995), p. 14.
278

67. 2:44 pm
Page 279
chapter 15 slow-onset disasters
ically into early-warning systems. Officials are more likely to be impressed by
conventional quantitative monitoring than community-based systems and
local knowledge.
Multi-indicator systems are sensitive to the complexity of famine processes,
and are therefore more likely to detect worsening food security early enough
for agencies to intervene to protect livelihood assets and prevent starvation.
However, information does not speak for itself. Data have to be interpreted.
Options for intervention must be assessed. This presents many challenges
where multiple sources of information are used.
The different types of data are not easily compared. For example, how does
one weigh up the relative significance of data on grain prices in local markets
compared to levels of rainfall or farm crop production, or sales of livelihood
assets? To add to the problem, most systems depend to some extent on proxy
indicators of food stress (e.g. the timing and extent of adoption of particular
coping strategies).
Different methods are needed to collect different kinds of information, each
requiring its own skills. Formal measuring systems can be used for some
aspects of food security, such as crop production and food prices, that are
quantifiable. Monitoring of nutritional status has its own methods. Assessing
wider household food security status requires expertise in interviewing and
participatory appraisal. These skills can be learnt and transferred, but this
takes time, and specialist assistance will be needed as it is unlikely that any
one local organisation or project team will have all of the relevant expertise
in-house. Rapid staff turnover often prevents skills from becoming fixed
within an organisation.
Some relevant information may have been collected by other people and for
other purposes (e.g. a Ministry of Agriculture will collect agricultural production data). It will have to be obtained from those users. This may not be easy,
especially in countries with very bureaucratic administrations. Information
from other sources may have been collected or aggregated on a different
basis from that of the local monitoring system. For example, government data
sets may gather information at village or even district level, rather than
household level. Government officials prefer to use administrative areas as
their units of analysis, and may not take account of geographical or social
differentiation within an administrative area. Disaggregation of data by age,
sex or occupation is likely to vary between different data sets, as will the
timing and frequency of data collection.
279
slow-onset
disasters
8/3/04
15
Good Practice 4th

68. Good Practice 4th
8/3/04
2:44 pm
Page 280
disaster risk reduction: mitigation and preparedness
Over-emphasis on data collection is a common failing. Information is often
gathered for its own sake, without sufficient thought being given to what field
agencies need to know. Local-level systems in particular are likely to find
themselves unable to process all the information they collect. This can be a
particular problem for those that use a wide range of data sources and indicators. There are several instances of projects having to scale down their datagathering operations, or bring them more into line with operational needs.
At the end of this process, information has to be packaged in a way that is
intelligible to decision-makers and that helps to guide them towards appropriate action. This link to action must be kept firmly in mind when planning
and running early-warning systems. The system may have to supply information to a wide variety of users, ranging from government policy-makers to
field managers. Each group may want different kinds of data, which may have
to be presented in different ways.
Case Study 15.8 illustrates one agency’s approach to collecting and analysing
food security information.
15.4.2 Maintaining local systems
Food security information systems are complex and difficult to manage. They
can also be costly because of the considerable staff time required in
collecting and interpreting data. This is true even with participatory datagathering methods that involve community members, because the information still has to be drawn together from different sources, analysed and then
packaged for decision-makers and field workers. As data often have to be
gathered from communities dispersed over a wide geographical area, transport and subsistence costs can also be high.
Owing to these factors, the sustainability of warning systems is a major challenge. Systems need to be maintained continuously to give reliable data of
patterns of food supply and demand over time. A secure funding stream is therefore needed. Lack of resources has damaged a number of government and NGOrun warning systems. The project-based approach that NGOs are generally
obliged to adopt is an insecure foundation for such work because of its fixed
time spans and the difficulty of obtaining repeat funding from donor agencies.
Both national and local warning systems must be integrated into the institutions that manage them. Many systems are purpose-built and tend to stand
alone. Those who set up early-warning systems should plan their external
linkages as carefully as their internal mechanisms. Local-level systems often
280

69. 8/3/04
2:44 pm
Page 281
15
chapter 15 slow-onset disasters
Case Study 15.8
Monitoring household food security
In the early 1990s, Save the Children
Fund (UK) adopted a ‘food economy’
approach to analysing household
food security. Food economy is
defined as the sum of ways families
obtain food. The food economy
approach is most commonly used to
estimate food aid needs, but it can
also be used to inform other kinds of
intervention to support food
production or livelihoods.
Central to this approach is
understanding rural households’
everyday circumstances. Knowing
how people normally obtain food is
an essential part of predicting how
they will react in a crisis, be it a
major disaster or a seasonal food
shortage. Building up a ‘normal year’
or baseline picture helps determine
key indicators of food security that
can then be monitored. Baseline
pictures contain information on how
households normally obtain food
and cash income, their connections
with the market and social or kinship
networks, their assets (land, food
stocks, livestock, cash, goods, tools)
and their expenditure patterns. Data
collection is mostly based on
fieldwork using RRA or PRA
techniques and, particularly, semistructured interviews with
community members, individually
and in groups.
The next steps are to identify potential
problems – changes in agricultural,
economic or security conditions – that
could affect access to food, and to
develop scenarios showing what the
impact of such changes would be. For
example, the impact of reduced crop
production, milk yields, and income
from livestock sales or wage labour
can be translated into an estimate of
the likely impact on food availability.
Similarly, the potential role of various
coping strategies can be estimated. A
computer software program, RiskMap,
has been developed for this scenario
analysis, which can be complex,
although for smaller data sets it can
be done using a spreadsheet or even
manually.
Data sets are grouped
geographically into ‘food economy
zones’: areas in which the same food
and cash income options are
available and relied upon to varying
degrees by families with different
levels of wealth. This makes it
possible to identify particular groups
in need, but data collected on the
basis of food economy zones are
often incompatible with other data
sets gathered on the basis of
administrative districts.
Owing to the methods used, much of
the information collected in the field
(continued)
281
slow-onset
disasters
Good Practice 4th

70. Good Practice 4th
8/3/04
2:44 pm
Page 282
disaster risk reduction: mitigation and preparedness
Case Study 15.8 (continued)
is, inevitably, quantitatively
imprecise. However, data analysis is
designed to point out
inconsistencies in data and ensure
that the overall picture adds up.
Field information of one kind is
checked against other kinds and
compared to secondary sources.
Analysis takes place in the field, so
that contradictions or odd findings
can be dealt with on the spot.
A major problem with this approach
is the need for skilled staff and
ongoing training. Well-educated and
committed field staff are not always
easy to find, and turnover can be
considerable.
Another problem is over-reliance on
standard reporting formats and
guidelines, leading to standardisation
of the information-gathering process
and hence to poor results. Field staff
need to be sensitive to local
circumstances and interviewees’
needs, and ready to adapt their datacollection techniques accordingly.
However, there is an ongoing tension
between this need for flexibility, and
the need for consistent data.
The food economy approach has
been used widely by Save the
Children Fund (UK) in Africa and
Asia, and both it and the RiskMap
programme have undergone
considerable development during
the past ten years. A manual has
been produced and field training
courses are held.
T. Boudreau, The Food Economy
Approach: A Framework for
Understanding Rural Livelihoods
(London: Humanitarian Practice Network,
Network Paper 26, 1998); J. Seaman,
‘Making Exchange Entitlements
Operational: The Food Economy Approach
to Famine Prediction and the RiskMap
Computer Program’, Disasters, vol. 24,
no. 2, 2000, pp. 133–52.
feed into national-level ones. But unless decentralised data are available for
all the areas at risk, this can distort decision-making by giving undue prominence to particular districts.
Other problems include the lack of integration between different agencies’
early-warning systems, which hinders collective analysis and action. Failure
to standardise data across systems is a major issue. Agencies fail to learn
lessons from each other’s experiences, and even from their own similar
programmes elsewhere.
282

71. 8/3/04
2:44 pm
Page 283
15
chapter 15 slow-onset disasters
Case Study 15.9
Sustaining food security information systems
In 1987, Save the Children (UK)
established a local food security
monitoring programme, Suivi
Alimentaire du Delta Seno (SADS), in
Mopti, Mali, in collaboration with
Oxfam (UK) and the International
Union for Conservation of Nature.
the misconception that food supply
alone determines the state of food
security. But coping strategies were
difficult to monitor because
households change their strategies
frequently, and good baseline
information was needed.
The system was designed to monitor
local food entitlements. Data were
collected using PRA methods. This
included information on agricultural
production, fish production,
livestock conditions, levels of onfarm stocks, off-farm employment,
household consumption and
migration. Food prices were also
monitored, together with the
marketing strategies of producers
and traders. Secondary data sets
containing statistical information
were incorporated: these covered
rainfall and flood levels and national
and regional food production
estimates. Several one-off surveys
were carried out to improve
knowledge of particular subjects,
such as the use of wild foods. The
system developed indicators of the
food security of different groups at
different times of year.
SADS produced quarterly food
security assessments for local,
national and regional decisionmakers in government and
international agencies. It also
intervened to support livelihood
strategies, principally through a
credit scheme introduced in 1991
that made loans for productive
activities such as the purchase of
seeds or fishing equipment, to
establish cereal banks and for
economic diversification.
The system’s focus on coping
strategies greatly improved
understanding of the complexity of
local food strategies, and challenged
The sustainability of the information
system and its institutional linkages
were not considered in depth at the
start of the project because the main
aim was to explore the nature of
vulnerability. It was difficult for an
NGO to maintain such a system,
owing to the cost of field staff and
data collection, and the need to train
people to gather and analyse
information. The information SADS
produced was timely in alerting
decision-makers to pockets of food
insecurity that had sometimes been
(continued)
283
slow-onset
disasters
Good Practice 4th

72. Good Practice 4th
8/3/04
2:44 pm
Page 284
disaster risk reduction: mitigation and preparedness
Case Study 15.9 (continued)
overlooked by national systems, but
its influence on decision-making was
limited, largely for institutional
reasons: it did not have direct links
with government.
For these and other reasons, the
information system was closed down
in 1993. It was felt that it would be
more appropriate to establish such a
scheme within local or regional
government structures; there was
already a national early-warning
system in Mali.
R. J. Lambert, ‘Monitoring Food Security
and Coping Strategies: Lessons from
Information Collection and Analysis in
Mopti, Mali’, Disasters, vol. 18, no. 4, 1994,
pp. 332–43.
15.5 Chapter summary
• There is a close link between environmental degradation and increased
risk from natural hazards. Environmental protection and renewal is
feasible, but may challenge vested interests.
• Industrial and domestic waste can present a serious threat to public
health, whose impact may not be felt until some time in the future.
• Significant reductions in pollution and improvements in waste management require the involvement of the state, but partnerships between
government and other actors are desirable.
• Global warming threatens to undermine society’s resilience to environmental stress. The best ways of coping with the resulting increase in
disaster risk are still unclear.
• Natural hazards such as drought are only one factor in food crises and
famines. Political, social and economic factors can be influential.
• Food security is not merely a question of food availability. People do not
have equal access to food because of differences in the resources they
possess and other economic, social and political factors.
• There are many ways of overcoming food insecurity by improving people’s
ability to obtain food. Such initiatives can be undertaken locally.
• The most effective way to protect communities against drought-induced
food insecurity and famine is to strengthen existing coping and livelihood
strategies well in advance. Yet most external interventions are still in
response to drought and crisis.
284

73. 2:44 pm
Page 285
chapter 15 slow-onset disasters
• The ideal approach is an integrated one that uses a range of methods to
tackle different aspects of food and related livelihood insecurity. This is
likely to include methods of protecting and increasing food production,
better preservation of food crops, water conservation, preserving livestock and broader livelihood support.
• Local food insecurity/famine early-warning systems are valuable, especially where they are based on a wide range of indicators and community
participation.
• Early-warning systems of all kinds need to stimulate appropriate action
early enough to prevent destitution and starvation. In practice, they may
fail to do so because of obstacles in the institutional and political environments in which they operate.
• The selection of food insecurity indicators and analysis of data need to be
planned and managed carefully to ensure the collection of appropriate
types and amounts of information.
• Food security information and early-warning systems are often costly,
complex and difficult to manage. Sustaining such systems can be a major
challenge.
• More harmonisation and integration of systems is needed.
Notes
1
2
3
4
5
6
W. Ahrens, ‘Cultural and Political Obstacles to Accepting Early Warnings of the Aral Sea
Disaster’, in J. Zschau and A. N. Küppers, Early Warning Systems for Natural Disaster Reduction
(Heidelberg: Springer Verlag, 2003), pp. 305–10.
K. Smith, Environmental Hazards: Assessing Risk and Reducing Disaster (London: Routledge,
1996), pp. 314–40.
W. N. Adger and N. Brooks, ‘Does Global Environmental Change Cause Vulnerability to
Disaster?’, in M. Pelling (ed.), Natural Disasters and Development in a Globalizing World
(London: Routledge, 2003), pp. 19–42.
S. Devereux, Famine in the Twentieth Century (Brighton: University of Sussex Institute of
Development Studies, Working Paper 105, 2000), www.ids.ac.uk/ids/bookshop/wp/wp105.pdf.
M. Hubbard, Improving Food Security: A Guide for Rural Development Managers (London: IT
Publications, 1995); J. Borton and N. Nicholds, Drought and Famine (Geneva/New York:
UNDP/DHA Disaster Management Training Programme, 1994),
http://undmtp.org/english/droughtandfamine_guide/drought_guide.pdf; H. Young et al.,
Food-security Assessments in Emergencies: A Livelihoods Approach (London: Humanitarian
Practice Network, Network Paper 36, 2001).
World Food Summit Plan of Action (Rome: UN Food and Agriculture Organization, 1996),
www.fao.org/wfs/index_en.htm.
285
slow-onset
disasters
8/3/04
15
Good Practice 4th

74. Good Practice 4th
8/3/04
2:44 pm
Page 286
disaster risk reduction: mitigation and preparedness
7 For drought, see Smith, Environmental Hazards, pp. 286–313; Borton and Nicholds, Drought
and Famine, pp. 11–22.
8 Borton and Nicholds, Drought and Famine, p. 15.
9 Young et al., Food-security Assessments in Emergencies, p. 4.
10 D. Cooper, R. Vellvé and H. Hobbelink (eds), Growing Diversity: Genetic Resources and Local
Food Security (London: IT Publications, 1992).
11 Hubbard, Improving Food Security, pp. 58–60; I. Carter, Improving Food Security: A PILLARS
Guide (London: TearFund, 2001); M. B. Anderson and P. J. Woodrow, Rising from the Ashes:
Development Strategies in Times of Disaster (London: IT Publications, 1998), pp. 185–205;
Developing Countries Farm Radio Network, www.farmradio.org.
12 Hubbard, Improving Food Security, pp. 90–98.
13 R. de. S. Ariyabandu, Wisdom of Traditional Collection of Rainwater for Domestic Use
(Colombo: Lanka Rainwater Harvesting Forum, 1998).
14 Hubbard, Improving Food Security, pp. 99–108.
15 M. R. Bhatt, ‘Gender and Disaster: Fodder Security System of Banaskantha Women’, mimeo
(Ahmedabad: Disaster Mitigation Institute, 1996).
16 For an introduction to the subject, see F. Mason and A. Taylor, ‘A Review of the Advances and
Challenges in Nutrition in Conflicts and Crises over the last 20 Years’, SCN News, no. 24, July
2002, pp: 21–27, www.unsystem.org/scn/Publications/SCNNews/scnnews24.pdf; Hubbard,
Improving Food Security, pp. 22–38.
17 Hubbard, Improving Food Security, pp. 39–62.
18 M. Buchanan-Smith and S. Davies, Famine Early Warning and Response – The Missing Link
(London: IT Publications, 1995).
19 Ibid., p. 202.
20 Ibid.; Borton and Nicholds, Drought and Famine; M. Omar, ‘Drought and Famine Crises
Management: A Training Programme for Voluntary Organisations in Sudan’, mimeo
(Shrivenham: Cranfield Disaster Preparedness Centre, 1995).
286